All_models.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "507130ba",
   "metadata": {},
   "source": [
    "**Training model on the raw data with 3 layers and data split**\n",
    "\n",
    "**Optimizer = Adam**\n",
    "\n",
    "**Loss = binary_crossentropy**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5b47133",
   "metadata": {},
   "source": [
    "# Folder Creation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d5637cc3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 1s 553ms/step\n",
      "1/1 [==============================] - 0s 418ms/step\n",
      "1/1 [==============================] - 0s 421ms/step\n",
      "1/1 [==============================] - 0s 471ms/step\n",
      "1/1 [==============================] - 0s 470ms/step\n",
      "1/1 [==============================] - 0s 484ms/step\n",
      "1/1 [==============================] - 0s 433ms/step\n",
      "1/1 [==============================] - 0s 456ms/step\n",
      "1/1 [==============================] - 0s 450ms/step\n",
      "1/1 [==============================] - 0s 452ms/step\n",
      "1/1 [==============================] - 0s 414ms/step\n",
      "1/1 [==============================] - 0s 429ms/step\n",
      "1/1 [==============================] - 0s 426ms/step\n",
      "1/1 [==============================] - 0s 439ms/step\n",
      "1/1 [==============================] - 0s 432ms/step\n",
      "1/1 [==============================] - 0s 432ms/step\n",
      "1/1 [==============================] - 0s 439ms/step\n",
      "1/1 [==============================] - 0s 450ms/step\n",
      "1/1 [==============================] - 0s 441ms/step\n",
      "1/1 [==============================] - 0s 423ms/step\n",
      "1/1 [==============================] - 0s 435ms/step\n",
      "1/1 [==============================] - 0s 412ms/step\n",
      "1/1 [==============================] - 0s 408ms/step\n",
      "1/1 [==============================] - 0s 408ms/step\n",
      "1/1 [==============================] - 0s 451ms/step\n",
      "1/1 [==============================] - 0s 431ms/step\n",
      "1/1 [==============================] - 0s 421ms/step\n",
      "1/1 [==============================] - 0s 420ms/step\n",
      "1/1 [==============================] - 0s 413ms/step\n",
      "1/1 [==============================] - 0s 418ms/step\n",
      "1/1 [==============================] - 0s 422ms/step\n",
      "1/1 [==============================] - 0s 420ms/step\n",
      "1/1 [==============================] - 0s 409ms/step\n",
      "1/1 [==============================] - 0s 430ms/step\n",
      "1/1 [==============================] - 0s 429ms/step\n",
      "1/1 [==============================] - 0s 425ms/step\n",
      "1/1 [==============================] - 0s 407ms/step\n",
      "1/1 [==============================] - 0s 415ms/step\n",
      "1/1 [==============================] - 0s 424ms/step\n",
      "1/1 [==============================] - 0s 453ms/step\n",
      "1/1 [==============================] - 0s 421ms/step\n",
      "1/1 [==============================] - 0s 420ms/step\n",
      "1/1 [==============================] - 0s 417ms/step\n",
      "1/1 [==============================] - 0s 425ms/step\n",
      "1/1 [==============================] - 0s 432ms/step\n",
      "1/1 [==============================] - 0s 419ms/step\n",
      "Healthy Images: 32\n",
      "Gumosis Images: 28\n",
      "Red Rust Images: 32\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing import image\n",
    "\n",
    "# Load the trained model\n",
    "model = tf.keras.models.load_model('F:/College/Years/B.Tech/Major_project/Model_resources/pest_and_insect_detection_model.h5')\n",
    "\n",
    "# Define the path to the folder containing test images\n",
    "test_image_dir = 'E:/Download/data/Categorized/Healthy'\n",
    "\n",
    "# Get a list of image file names in the test folder\n",
    "image_files = [f for f in os.listdir(test_image_dir) if os.path.isfile(os.path.join(test_image_dir, f))]\n",
    "\n",
    "# Define a dictionary to map class indices to disease names\n",
    "class_names = {0: 'Gumosis', 1: 'Healthy', 2: 'Red Rust'}\n",
    "\n",
    "# Define the remedies dictionary\n",
    "remedies = {\n",
    "    \"Gumosis\": {\n",
    "        \"fungicides\": [\"Bordeaux mixture (1%)\", \"Copper oxychloride (0.3%)\", \"Mancozeb (0.3%)\"],\n",
    "        \"other_treatments\": [\"Remove and destroy affected branches.\", \"Apply a copper-based fungicide to the cut surfaces.\", \"Apply a balanced fertilizer to the tree.\", \"Water the tree regularly, but avoid overwatering.\"]\n",
    "    },\n",
    "    \"Healthy\": {\n",
    "        \"preventive_measures\": [\"Plant cashew trees in well-drained soil.\", \"Avoid overwatering the trees.\", \"Apply a balanced fertilizer to the trees regularly.\", \"Inspect the trees regularly for pests and diseases.\"],\n",
    "        \"treatment_of_pests_and_diseases\": [\"Use fungicides and pesticides as needed.\", \"Remove and destroy affected branches.\", \"Apply copper-based fungicides to the cut surfaces.\"]\n",
    "    },\n",
    "    \"Red Rust\": {\n",
    "        \"fungicides\": [\"Bordeaux mixture (1%)\", \"Copper oxychloride (0.3%)\", \"Mancozeb (0.3%)\"],\n",
    "        \"other_treatments\": [\"Remove and destroy affected leaves.\", \"Spray the tree with a copper-based fungicide.\", \"Apply a balanced fertilizer to the tree.\", \"Water the tree regularly, but avoid overwatering.\"]\n",
    "    }\n",
    "}\n",
    "\n",
    "# Iterate through the test images and make predictions\n",
    "for image_file in image_files:\n",
    "    try:\n",
    "        # Load and preprocess the test image\n",
    "        img_path = os.path.join(test_image_dir, image_file)\n",
    "        img = image.load_img(img_path, target_size=(400, 400))\n",
    "        img_array = image.img_to_array(img)\n",
    "        img_array = np.expand_dims(img_array, axis=0)\n",
    "        img_array /= 255.0  # Rescale pixel values to [0, 1]\n",
    "\n",
    "        # Make predictions\n",
    "        predictions = model.predict(img_array)\n",
    "        predicted_class = np.argmax(predictions, axis=1)[0]\n",
    "\n",
    "        # Get the disease name from the dictionary\n",
    "        detected_disease = class_names[predicted_class]\n",
    "\n",
    "        # Get the remedies for the detected disease\n",
    "        disease_remedies = remedies.get(detected_disease, {})\n",
    "        fungicides = disease_remedies.get(\"fungicides\", [])\n",
    "        other_treatments = disease_remedies.get(\"other_treatments\", [])\n",
    "\n",
    "        # Create the destination folder if it doesn't exist\n",
    "        destination_folder = os.path.join(test_image_dir, detected_disease)\n",
    "        if not os.path.exists(destination_folder):\n",
    "            os.makedirs(destination_folder)\n",
    "\n",
    "        # Move the image to the respective disease category folder\n",
    "        os.rename(img_path, os.path.join(destination_folder, image_file))\n",
    "\n",
    "        # Create a text file with remedy information\n",
    "        with open(os.path.join(destination_folder, f\"{os.path.splitext(image_file)[0]}.txt\"), 'w') as remedy_file:\n",
    "            remedy_file.write(f\"Detected Disease: {detected_disease}\\n\\nFungicides:\\n{', '.join(fungicides)}\\n\\nOther Treatments:\\n{', '.join(other_treatments)}\")\n",
    "\n",
    "    except Exception as e:\n",
    "        print(f\"Error processing {image_file}: {str(e)}\")\n",
    "\n",
    "# Print the counts of healthy and disease images\n",
    "print(f\"Healthy Images: {len(os.listdir(os.path.join(test_image_dir, 'Healthy')))}\")\n",
    "print(f\"Gumosis Images: {len(os.listdir(os.path.join(test_image_dir, 'Gumosis')))}\")\n",
    "print(f\"Red Rust Images: {len(os.listdir(os.path.join(test_image_dir, 'Red Rust')))}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "938739bf",
   "metadata": {},
   "source": [
    "**Fine tunining the model**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82b161bf",
   "metadata": {},
   "source": [
    "# **Model training (Cashew)(best accuracy)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "dec75eb4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 2755 images belonging to 3 classes.\n",
      "Found 687 images belonging to 3 classes.\n",
      "Epoch 1/10\n",
      "86/86 [==============================] - 1200s 14s/step - loss: 1.8112 - accuracy: 0.7822 - val_loss: 0.1952 - val_accuracy: 0.9405\n",
      "Epoch 2/10\n",
      "86/86 [==============================] - 1237s 14s/step - loss: 0.1732 - accuracy: 0.9372 - val_loss: 0.1488 - val_accuracy: 0.9509\n",
      "Epoch 3/10\n",
      "86/86 [==============================] - 1230s 14s/step - loss: 0.1839 - accuracy: 0.9317 - val_loss: 0.1107 - val_accuracy: 0.9643\n",
      "Epoch 4/10\n",
      "86/86 [==============================] - 1228s 14s/step - loss: 0.1655 - accuracy: 0.9346 - val_loss: 0.0874 - val_accuracy: 0.9747\n",
      "Epoch 5/10\n",
      "86/86 [==============================] - 1175s 14s/step - loss: 0.1592 - accuracy: 0.9416 - val_loss: 0.1116 - val_accuracy: 0.9598\n",
      "Epoch 6/10\n",
      "86/86 [==============================] - 1181s 14s/step - loss: 0.1218 - accuracy: 0.9603 - val_loss: 0.0600 - val_accuracy: 0.9807\n",
      "Epoch 7/10\n",
      "86/86 [==============================] - 1173s 14s/step - loss: 0.1143 - accuracy: 0.9567 - val_loss: 0.1308 - val_accuracy: 0.9494\n",
      "Epoch 8/10\n",
      "86/86 [==============================] - 1170s 14s/step - loss: 0.1309 - accuracy: 0.9530 - val_loss: 0.0907 - val_accuracy: 0.9673\n",
      "Epoch 9/10\n",
      "86/86 [==============================] - 1169s 14s/step - loss: 0.1296 - accuracy: 0.9508 - val_loss: 0.1331 - val_accuracy: 0.9568\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ABHIJIT DESHPANDE\\anaconda3\\lib\\site-packages\\keras\\src\\engine\\training.py:3000: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
      "  saving_api.save_model(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 687 images belonging to 3 classes.\n",
      "21/21 [==============================] - 231s 11s/step - loss: 0.0502 - accuracy: 0.9807\n",
      "Testing Accuracy: 0.980654776096344\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.applications import VGG16\n",
    "from tensorflow.keras import models, layers, optimizers\n",
    "\n",
    "# Set the image dimensions and other parameters\n",
    "img_width, img_height = 400, 400\n",
    "batch_size = 32\n",
    "epochs = 10\n",
    "num_classes = 3\n",
    "\n",
    "# Set the data directory\n",
    "data_dir = 'F:\\College\\Years\\B.Tech\\Major_project\\Cashew'\n",
    "\n",
    "# Data augmentation and other preprocessing for the training dataset\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1.0 / 255,\n",
    "    shear_range=0.2,\n",
    "    zoom_range=0.2,\n",
    "    horizontal_flip=True,\n",
    "    rotation_range=20,       # Additional data augmentation: rotation\n",
    "    width_shift_range=0.2,   # Additional data augmentation: horizontal shift\n",
    "    height_shift_range=0.2,  # Additional data augmentation: vertical shift\n",
    "    validation_split=0.2     # Split data into training and validation sets\n",
    ")\n",
    "\n",
    "# Load the training and validation datasets with data augmentation\n",
    "train_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='training'\n",
    ")\n",
    "\n",
    "validation_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation'\n",
    ")\n",
    "\n",
    "# Load the pre-trained VGG16 model without the top (fully connected) layers\n",
    "base_model = VGG16(weights='imagenet', include_top=False, input_shape=(img_width, img_height, 3))\n",
    "\n",
    "# Freeze the layers of the pre-trained VGG16 base to avoid overfitting\n",
    "for layer in base_model.layers:\n",
    "    layer.trainable = False\n",
    "\n",
    "# Add new layers for our specific classification task\n",
    "model = models.Sequential()\n",
    "model.add(base_model)\n",
    "model.add(layers.Flatten())\n",
    "model.add(layers.Dense(256, activation='relu'))\n",
    "model.add(layers.Dropout(0.5))\n",
    "model.add(layers.Dense(num_classes, activation='softmax'))\n",
    "\n",
    "# Compile the model\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "# Train the model with early stopping\n",
    "early_stopping = tf.keras.callbacks.EarlyStopping(patience=3, restore_best_weights=True)\n",
    "history = model.fit(\n",
    "    train_generator,\n",
    "    steps_per_epoch=train_generator.samples // batch_size,\n",
    "    validation_data=validation_generator,\n",
    "    validation_steps=validation_generator.samples // batch_size,\n",
    "    epochs=epochs,\n",
    "    callbacks=[early_stopping]\n",
    ")\n",
    "\n",
    "# Save the model\n",
    "model.save('pest_and_insect_detection_model.h5')\n",
    "\n",
    "# Evaluate the model on the testing set\n",
    "test_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation',\n",
    "    shuffle=False\n",
    ")\n",
    "\n",
    "loss, accuracy = model.evaluate(\n",
    "    test_generator,\n",
    "    steps=test_generator.samples // batch_size\n",
    ")\n",
    "\n",
    "print(f\"Testing Accuracy: {accuracy}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f1d75c3",
   "metadata": {},
   "source": [
    "# Detected"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f0af716e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ABHIJIT DESHPANDE\\anaconda3\\lib\\site-packages\\scipy\\__init__.py:146: UserWarning: A NumPy version >=1.16.5 and <1.23.0 is required for this version of SciPy (detected version 1.24.3\n",
      "  warnings.warn(f\"A NumPy version >={np_minversion} and <{np_maxversion}\"\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing import image\n",
    "\n",
    "# Load the trained model\n",
    "model = tf.keras.models.load_model('F:/College/Years/B.Tech/Major_project/Model_resources/pest_and_insect_detection_model.h5')\n",
    "\n",
    "# Define the path to the folder containing test images\n",
    "test_image_dir = 'E:/Download/data/G'\n",
    "\n",
    "# Get a list of image file names in the test folder\n",
    "image_files = [f for f in os.listdir(test_image_dir) if os.path.isfile(os.path.join(test_image_dir, f))]\n",
    "\n",
    "# Define a dictionary to map class indices to disease names\n",
    "class_names = {0: 'Gumosis', 1: 'Healthy', 2: 'Red Rust'}\n",
    "\n",
    "# Iterate through the test images and make predictions\n",
    "for image_file in image_files:\n",
    "    try:\n",
    "        # Load and preprocess the test image\n",
    "        img_path = os.path.join(test_image_dir, image_file)\n",
    "        img = image.load_img(img_path, target_size=(400, 400))\n",
    "        img_array = image.img_to_array(img)\n",
    "        img_array = np.expand_dims(img_array, axis=0)\n",
    "        img_array /= 255.0  # Rescale pixel values to [0, 1]\n",
    "\n",
    "        # Make predictions\n",
    "        predictions = model.predict(img_array)\n",
    "        predicted_class = np.argmax(predictions, axis=1)[0]\n",
    "\n",
    "        # Get the disease name from the dictionary\n",
    "        detected_disease = class_names[predicted_class]\n",
    "\n",
    "        # Print the result for each image\n",
    "        print(f\"Image: {image_file}, Detected Disease: {detected_disease}\")\n",
    "    \n",
    "    except Exception as e:\n",
    "        print(f\"Error processing {image_file}: {str(e)}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2257292b",
   "metadata": {},
   "source": [
    "# Casava training and testing Final \n",
    "bacteria, green, brown, health, mosaic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "69cd8d3d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 4878 images belonging to 4 classes.\n",
      "Found 1217 images belonging to 4 classes.\n",
      "Epoch 1/50\n",
      "152/152 [==============================] - 164s 1s/step - loss: 0.7391 - accuracy: 0.7134 - val_loss: 0.5325 - val_accuracy: 0.7928\n",
      "Epoch 2/50\n",
      "152/152 [==============================] - 114s 750ms/step - loss: 0.5024 - accuracy: 0.8062 - val_loss: 0.5923 - val_accuracy: 0.7484\n",
      "Epoch 3/50\n",
      "152/152 [==============================] - 113s 745ms/step - loss: 0.4553 - accuracy: 0.8203 - val_loss: 0.5142 - val_accuracy: 0.7862\n",
      "Epoch 4/50\n",
      "152/152 [==============================] - 113s 743ms/step - loss: 0.4175 - accuracy: 0.8405 - val_loss: 0.4837 - val_accuracy: 0.8109\n",
      "Epoch 5/50\n",
      "152/152 [==============================] - 113s 741ms/step - loss: 0.4051 - accuracy: 0.8487 - val_loss: 0.4672 - val_accuracy: 0.8191\n",
      "Epoch 6/50\n",
      "152/152 [==============================] - 118s 776ms/step - loss: 0.3998 - accuracy: 0.8434 - val_loss: 0.4676 - val_accuracy: 0.8166\n",
      "Epoch 7/50\n",
      "152/152 [==============================] - 124s 818ms/step - loss: 0.3972 - accuracy: 0.8452 - val_loss: 0.4414 - val_accuracy: 0.8248\n",
      "Epoch 8/50\n",
      "152/152 [==============================] - 129s 850ms/step - loss: 0.3897 - accuracy: 0.8564 - val_loss: 0.5143 - val_accuracy: 0.8035\n",
      "Epoch 9/50\n",
      "152/152 [==============================] - 124s 817ms/step - loss: 0.3757 - accuracy: 0.8533 - val_loss: 0.4548 - val_accuracy: 0.8240\n",
      "Epoch 10/50\n",
      "152/152 [==============================] - 124s 817ms/step - loss: 0.3649 - accuracy: 0.8611 - val_loss: 0.4547 - val_accuracy: 0.8215\n",
      "Epoch 11/50\n",
      "152/152 [==============================] - 122s 804ms/step - loss: 0.3486 - accuracy: 0.8667 - val_loss: 0.4694 - val_accuracy: 0.8207\n",
      "Epoch 12/50\n",
      "152/152 [==============================] - 113s 741ms/step - loss: 0.3515 - accuracy: 0.8655 - val_loss: 0.4681 - val_accuracy: 0.8265\n",
      "Epoch 13/50\n",
      "152/152 [==============================] - 117s 769ms/step - loss: 0.3514 - accuracy: 0.8601 - val_loss: 0.4283 - val_accuracy: 0.8215\n",
      "Epoch 14/50\n",
      "152/152 [==============================] - 111s 730ms/step - loss: 0.3493 - accuracy: 0.8673 - val_loss: 0.4609 - val_accuracy: 0.8166\n",
      "Epoch 15/50\n",
      "152/152 [==============================] - 111s 728ms/step - loss: 0.3319 - accuracy: 0.8706 - val_loss: 0.4215 - val_accuracy: 0.8289\n",
      "Epoch 16/50\n",
      "152/152 [==============================] - 112s 739ms/step - loss: 0.3380 - accuracy: 0.8681 - val_loss: 0.4534 - val_accuracy: 0.8257\n",
      "Epoch 17/50\n",
      "152/152 [==============================] - 112s 735ms/step - loss: 0.3378 - accuracy: 0.8708 - val_loss: 0.4720 - val_accuracy: 0.8067\n",
      "Epoch 18/50\n",
      "152/152 [==============================] - 115s 755ms/step - loss: 0.3182 - accuracy: 0.8760 - val_loss: 0.4468 - val_accuracy: 0.8224\n",
      "Epoch 19/50\n",
      "152/152 [==============================] - 115s 760ms/step - loss: 0.3347 - accuracy: 0.8686 - val_loss: 0.4474 - val_accuracy: 0.8314\n",
      "Epoch 20/50\n",
      "152/152 [==============================] - 115s 755ms/step - loss: 0.3093 - accuracy: 0.8754 - val_loss: 0.4132 - val_accuracy: 0.8495\n",
      "Epoch 21/50\n",
      "152/152 [==============================] - 114s 750ms/step - loss: 0.3169 - accuracy: 0.8762 - val_loss: 0.4675 - val_accuracy: 0.8183\n",
      "Epoch 22/50\n",
      "152/152 [==============================] - 114s 751ms/step - loss: 0.3201 - accuracy: 0.8737 - val_loss: 0.4150 - val_accuracy: 0.8454\n",
      "Epoch 23/50\n",
      "152/152 [==============================] - 115s 752ms/step - loss: 0.3194 - accuracy: 0.8762 - val_loss: 0.4559 - val_accuracy: 0.8363\n",
      "Epoch 24/50\n",
      "152/152 [==============================] - 120s 789ms/step - loss: 0.2966 - accuracy: 0.8809 - val_loss: 0.5183 - val_accuracy: 0.8059\n",
      "Epoch 25/50\n",
      "152/152 [==============================] - 113s 741ms/step - loss: 0.3064 - accuracy: 0.8846 - val_loss: 0.5233 - val_accuracy: 0.8166\n",
      "Epoch 26/50\n",
      "152/152 [==============================] - 114s 753ms/step - loss: 0.3155 - accuracy: 0.8772 - val_loss: 0.4998 - val_accuracy: 0.8240\n",
      "Epoch 27/50\n",
      "152/152 [==============================] - 114s 752ms/step - loss: 0.3116 - accuracy: 0.8778 - val_loss: 0.4790 - val_accuracy: 0.8125\n",
      "Epoch 28/50\n",
      "152/152 [==============================] - 112s 735ms/step - loss: 0.2949 - accuracy: 0.8890 - val_loss: 0.4720 - val_accuracy: 0.8339\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ABHIJIT DESHPANDE\\anaconda3\\lib\\site-packages\\keras\\src\\engine\\training.py:3000: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
      "  saving_api.save_model(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1217 images belonging to 4 classes.\n",
      "38/38 [==============================] - 22s 584ms/step - loss: 0.4302 - accuracy: 0.8380\n",
      "Testing Accuracy: 0.8379934430122375\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.applications import MobileNetV2\n",
    "from tensorflow.keras import models, layers, optimizers\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# Set the image dimensions and other parameters\n",
    "img_width, img_height = 224, 224\n",
    "batch_size = 32\n",
    "epochs = 50\n",
    "num_classes = 4\n",
    "\n",
    "# Set the data directory\n",
    "data_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Raw_Data/CCMT_Dataset/Cassava'\n",
    "\n",
    "# Data augmentation and preprocessing for the training dataset\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1.0 / 255,\n",
    "    shear_range=0.2,\n",
    "    zoom_range=0.2,\n",
    "    horizontal_flip=True,\n",
    "    rotation_range=20,\n",
    "    width_shift_range=0.3,\n",
    "    height_shift_range=0.2,\n",
    "    validation_split=0.2\n",
    ")\n",
    "\n",
    "# Load the data and split into training and validation sets\n",
    "data_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='training'\n",
    ")\n",
    "\n",
    "validation_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation'\n",
    ")\n",
    "\n",
    "# Load the MobileNetV2 model without the top (fully connected) layers\n",
    "base_model = tf.keras.applications.MobileNetV2(weights='imagenet', include_top=False, input_shape=(img_width, img_height, 3))\n",
    "\n",
    "# Freeze the layers of the base MobileNetV2 model\n",
    "for layer in base_model.layers:\n",
    "    layer.trainable = False\n",
    "\n",
    "# Add new layers for classification\n",
    "model = models.Sequential()\n",
    "model.add(base_model)\n",
    "model.add(layers.GlobalAveragePooling2D())\n",
    "model.add(layers.Dense(256, activation='relu'))\n",
    "model.add(layers.Dropout(0.5))\n",
    "model.add(layers.Dense(num_classes, activation='softmax'))\n",
    "\n",
    "# Compile the model\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "# Train the model with early stopping\n",
    "early_stopping = tf.keras.callbacks.EarlyStopping(patience=8, restore_best_weights=True)\n",
    "history = model.fit(\n",
    "    data_generator,\n",
    "    steps_per_epoch=data_generator.samples // batch_size,\n",
    "    validation_data=validation_generator,\n",
    "    validation_steps=validation_generator.samples // batch_size,\n",
    "    epochs=epochs,\n",
    "    callbacks=[early_stopping]\n",
    ")\n",
    "\n",
    "# Save the model\n",
    "model.save('Cassava_MobileNet_model.h5')\n",
    "\n",
    "# Evaluate the model on the testing set\n",
    "test_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation',\n",
    "    shuffle=False\n",
    ")\n",
    "\n",
    "loss, accuracy = model.evaluate(\n",
    "    test_generator,\n",
    "    steps=test_generator.samples // batch_size\n",
    ")\n",
    "\n",
    "print(f\"Testing Accuracy: {accuracy}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "6c64554e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 1s 633ms/step\n",
      "Image: bacterial blight10_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: bacterial blight11_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: bacterial blight12_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: bacterial blight13_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: bacterial blight14_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "Image: bacterial blight15_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: bacterial blight16_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: bacterial blight17_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: bacterial blight18_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "Image: bacterial blight19_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: bacterial blight20_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: bacterial blight21_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: bacterial blight22_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: bacterial blight23_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "Image: bacterial blight24_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: bacterial blight25_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: bacterial blight26_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: bacterial blight27_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: bacterial blight28_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: bacterial blight29_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: bacterial blight2_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: bacterial blight30_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: bacterial blight31_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 54ms/step\n",
      "Image: bacterial blight32_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: bacterial blight33_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: bacterial blight34_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: bacterial blight35_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: bacterial blight36_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: bacterial blight37_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: bacterial blight38_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: bacterial blight39_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: bacterial blight3_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: bacterial blight40_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: bacterial blight41_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: bacterial blight42_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: bacterial blight43_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: bacterial blight44_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: bacterial blight45_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 57ms/step\n",
      "Image: bacterial blight46_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: bacterial blight47_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: bacterial blight48_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: bacterial blight49_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: bacterial blight4_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: bacterial blight50_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: bacterial blight51_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 56ms/step\n",
      "Image: bacterial blight52_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: bacterial blight53_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: bacterial blight54_.jpg, Detected Disease: mosaic\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: bacterial blight55_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 58ms/step\n",
      "Image: bacterial blight56_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "Image: bacterial blight57_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: bacterial blight58_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: bacterial blight59_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: bacterial blight5_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: bacterial blight60_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: bacterial blight6_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: bacterial blight7_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 57ms/step\n",
      "Image: bacterial blight8_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: bacterial blight9_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: brown spot10_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot11_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: brown spot12_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 58ms/step\n",
      "Image: brown spot13_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: brown spot14_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: brown spot15_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "Image: brown spot16_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot17_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot18_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: brown spot19_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: brown spot20_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: brown spot21_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot22_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: brown spot23_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: brown spot24_.jpg, Detected Disease: bacterial blight\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot25_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "Image: brown spot26_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: brown spot27_.jpg, Detected Disease: mosaic\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: brown spot28_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: brown spot29_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: brown spot2_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: brown spot30_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: brown spot31_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 55ms/step\n",
      "Image: brown spot32_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: brown spot33_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: brown spot34_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: brown spot35_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot36_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: brown spot37_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: brown spot38_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: brown spot39_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 55ms/step\n",
      "Image: brown spot3_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: brown spot40_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot41_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: brown spot42_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: brown spot43_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: brown spot44_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "Image: brown spot45_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "Image: brown spot46_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: brown spot47_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: brown spot48_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: brown spot49_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: brown spot4_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: brown spot50_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: brown spot51_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "Image: brown spot52_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: brown spot53_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: brown spot54_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "Image: brown spot55_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: brown spot56_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: brown spot57_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot58_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: brown spot5_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot6_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: brown spot7_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: brown spot8_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "Image: brown spot9_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: green mite10_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: green mite11_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "Image: green mite12_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: green mite13_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: green mite14_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "Image: green mite15_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: green mite16_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: green mite17_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: green mite18_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite19_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: green mite20_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: green mite21_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: green mite22_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite23_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: green mite24_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: green mite25_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: green mite26_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: green mite27_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: green mite28_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: green mite29_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: green mite2_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite30_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "Image: green mite31_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: green mite32_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: green mite33_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: green mite34_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: green mite35_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: green mite36_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: green mite37_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: green mite38_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: green mite39_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: green mite3_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite40_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite41_.jpg, Detected Disease: bacterial blight\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: green mite42_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: green mite43_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: green mite44_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: green mite45_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: green mite46_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: green mite47_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "Image: green mite48_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "Image: green mite49_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: green mite4_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite50_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: green mite51_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: green mite52_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: green mite53_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: green mite54_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: green mite55_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite56_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite57_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: green mite5_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: green mite6_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: green mite7_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 60ms/step\n",
      "Image: green mite8_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: green mite9_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "Image: healthy10_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: healthy11_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 39ms/step\n",
      "Image: healthy12_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: healthy13_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: healthy14_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: healthy15_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: healthy16_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: healthy17_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy18_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: healthy19_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "Image: healthy20_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy21_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 39ms/step\n",
      "Image: healthy22_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "Image: healthy23_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: healthy24_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: healthy25_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: healthy26_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: healthy27_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy28_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy29_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: healthy2_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "Image: healthy30_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: healthy31_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: healthy32_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: healthy33_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: healthy34_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: healthy35_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: healthy36_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: healthy37_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy38_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: healthy39_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: healthy3_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: healthy40_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "Image: healthy41_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: healthy42_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: healthy43_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: healthy44_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy45_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: healthy46_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: healthy47_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy48_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy49_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: healthy4_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: healthy50_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "Image: healthy51_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: healthy52_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: healthy53_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: healthy54_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: healthy55_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy56_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy57_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy5_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: healthy6_.jpg, Detected Disease: bacterial blight\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: healthy7_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: healthy8_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: healthy9_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "Image: mosaic10_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "Image: mosaic11_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: mosaic12_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: mosaic13_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: mosaic20_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: mosaic21_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: mosaic22_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: mosaic23_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "Image: mosaic24_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: mosaic25_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 42ms/step\n",
      "Image: mosaic26_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "Image: mosaic27_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "Image: mosaic28_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: mosaic29_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: mosaic2_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: mosaic30_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "Image: mosaic31_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: mosaic32_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: mosaic33_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: mosaic34_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "Image: mosaic35_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: mosaic36_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: mosaic37_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: mosaic38_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "Image: mosaic39_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: mosaic3_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "Image: mosaic40_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: mosaic41_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: mosaic42_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: mosaic43_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: mosaic44_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: mosaic45_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "Image: mosaic46_.jpg, Detected Disease: Green mit\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: mosaic47_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 59ms/step\n",
      "Image: mosaic48_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: mosaic49_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: mosaic4_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: mosaic50_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: mosaic51_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "Image: mosaic52_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: mosaic53_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: mosaic54_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "Image: mosaic55_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "Image: mosaic56_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 52ms/step\n",
      "Image: mosaic57_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 51ms/step\n",
      "Image: mosaic58_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 43ms/step\n",
      "Image: mosaic59_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 48ms/step\n",
      "Image: mosaic5_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 55ms/step\n",
      "Image: mosaic60_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 58ms/step\n",
      "Image: mosaic61_.jpg, Detected Disease: Brown Spot\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "Image: mosaic62_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 49ms/step\n",
      "Image: mosaic63_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "Image: mosaic6_.jpg, Detected Disease: healthy\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: mosaic7_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: mosaic8_.jpg, Detected Disease: bacterial blight\n",
      "1/1 [==============================] - 0s 46ms/step\n",
      "Image: mosaic9_.jpg, Detected Disease: healthy\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing import image\n",
    "\n",
    "# Load the trained model\n",
    "model = tf.keras.models.load_model('MobileNet_model.h5')\n",
    "\n",
    "# Define the path to the folder containing test images\n",
    "test_image_dir = 'E:/Download/data/C'\n",
    "\n",
    "# Get a list of image file names in the test folder\n",
    "image_files = [f for f in os.listdir(test_image_dir) if os.path.isfile(os.path.join(test_image_dir, f))]\n",
    "\n",
    "# Define a dictionary to map class indices to disease names\n",
    "class_names = {0: 'bacterial blight', 1: 'Brown Spot', 2: 'Green mit', 3:'healthy', 4:'mosaic'}\n",
    "\n",
    "# Iterate through the test images and make predictions\n",
    "for image_file in image_files:\n",
    "    try:\n",
    "        # Load and preprocess the test image\n",
    "        img_path = os.path.join(test_image_dir, image_file)\n",
    "        img = image.load_img(img_path, target_size=(224, 224))\n",
    "        img_array = image.img_to_array(img)\n",
    "        img_array = np.expand_dims(img_array, axis=0)\n",
    "        img_array /= 255.0  # Rescale pixel values to [0, 1]\n",
    "\n",
    "        # Make predictions\n",
    "        predictions = model.predict(img_array)\n",
    "        predicted_class = np.argmax(predictions, axis=1)[0]\n",
    "\n",
    "        # Get the disease name from the dictionary\n",
    "        detected_disease = class_names[predicted_class]\n",
    "\n",
    "        # Print the result for each image\n",
    "        print(f\"Image: {image_file}, Detected Disease: {detected_disease}\")\n",
    "    \n",
    "    except Exception as e:\n",
    "        print(f\"Error processing {image_file}: {str(e)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "ffd00216",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resized mosaic1000_.jpg to (224, 224)\n",
      "Resized mosaic1001_.jpg to (224, 224)\n",
      "Resized mosaic1002_.jpg to (224, 224)\n",
      "Resized mosaic1003_.jpg to (224, 224)\n",
      "Resized mosaic1004_.jpg to (224, 224)\n",
      "Resized mosaic1005_.jpg to (224, 224)\n",
      "Resized mosaic1006_.jpg to (224, 224)\n",
      "Resized mosaic1007_.jpg to (224, 224)\n",
      "Resized mosaic1008_.jpg to (224, 224)\n",
      "Resized mosaic1009_.jpg to (224, 224)\n",
      "Resized mosaic100_.jpg to (224, 224)\n",
      "Resized mosaic1010_.jpg to (224, 224)\n",
      "Resized mosaic1011_.jpg to (224, 224)\n",
      "Resized mosaic1012_.jpg to (224, 224)\n",
      "Resized mosaic1013_.jpg to (224, 224)\n",
      "Resized mosaic1014_.jpg to (224, 224)\n",
      "Resized mosaic1015_.jpg to (224, 224)\n",
      "Resized mosaic1016_.jpg to (224, 224)\n",
      "Resized mosaic1017_.jpg to (224, 224)\n",
      "Resized mosaic1018_.jpg to (224, 224)\n",
      "Resized mosaic1019_.jpg to (224, 224)\n",
      "Resized mosaic101_.jpg to (224, 224)\n",
      "Resized mosaic1020_.jpg to (224, 224)\n",
      "Resized mosaic1021_.jpg to (224, 224)\n",
      "Resized mosaic1022_.jpg to (224, 224)\n",
      "Resized mosaic1023_.jpg to (224, 224)\n",
      "Resized mosaic1024_.jpg to (224, 224)\n",
      "Resized mosaic1025_.jpg to (224, 224)\n",
      "Resized mosaic1026_.jpg to (224, 224)\n",
      "Resized mosaic1027_.jpg to (224, 224)\n",
      "Resized mosaic1028_.jpg to (224, 224)\n",
      "Resized mosaic1029_.jpg to (224, 224)\n",
      "Resized mosaic102_.jpg to (224, 224)\n",
      "Resized mosaic1030_.jpg to (224, 224)\n",
      "Resized mosaic1031_.jpg to (224, 224)\n",
      "Resized mosaic1032_.jpg to (224, 224)\n",
      "Resized mosaic1033_.jpg to (224, 224)\n",
      "Resized mosaic1034_.jpg to (224, 224)\n",
      "Resized mosaic1035_.jpg to (224, 224)\n",
      "Resized mosaic1036_.jpg to (224, 224)\n",
      "Resized mosaic1037_.jpg to (224, 224)\n",
      "Resized mosaic1038_.jpg to (224, 224)\n",
      "Resized mosaic1039_.jpg to (224, 224)\n",
      "Resized mosaic103_.jpg to (224, 224)\n",
      "Resized mosaic1040_.jpg to (224, 224)\n",
      "Resized mosaic1041_.jpg to (224, 224)\n",
      "Resized mosaic1042_.jpg to (224, 224)\n",
      "Resized mosaic1043_.jpg to (224, 224)\n",
      "Resized mosaic1044_.jpg to (224, 224)\n",
      "Resized mosaic1045_.jpg to (224, 224)\n",
      "Resized mosaic1046_.jpg to (224, 224)\n",
      "Resized mosaic1047_.jpg to (224, 224)\n",
      "Resized mosaic1048_.jpg to (224, 224)\n",
      "Resized mosaic1049_.jpg to (224, 224)\n",
      "Resized mosaic104_.jpg to (224, 224)\n",
      "Resized mosaic1050_.jpg to (224, 224)\n",
      "Resized mosaic1051_.jpg to (224, 224)\n",
      "Resized mosaic1052_.jpg to (224, 224)\n",
      "Resized mosaic1053_.jpg to (224, 224)\n",
      "Resized mosaic1054_.jpg to (224, 224)\n",
      "Resized mosaic1055_.jpg to (224, 224)\n",
      "Resized mosaic1056_.jpg to (224, 224)\n",
      "Resized mosaic1057_.jpg to (224, 224)\n",
      "Resized mosaic1058_.jpg to (224, 224)\n",
      "Resized mosaic1059_.jpg to (224, 224)\n",
      "Resized mosaic105_.jpg to (224, 224)\n",
      "Resized mosaic1060_.jpg to (224, 224)\n",
      "Resized mosaic1061_.jpg to (224, 224)\n",
      "Resized mosaic1062_.jpg to (224, 224)\n",
      "Resized mosaic1063_.jpg to (224, 224)\n",
      "Resized mosaic1064_.jpg to (224, 224)\n",
      "Resized mosaic1065_.jpg to (224, 224)\n",
      "Resized mosaic1066_.jpg to (224, 224)\n",
      "Resized mosaic1067_.jpg to (224, 224)\n",
      "Resized mosaic1068_.jpg to (224, 224)\n",
      "Resized mosaic1069_.jpg to (224, 224)\n",
      "Resized mosaic106_.jpg to (224, 224)\n",
      "Resized mosaic1070_.jpg to (224, 224)\n",
      "Resized mosaic1071_.jpg to (224, 224)\n",
      "Resized mosaic1072_.jpg to (224, 224)\n",
      "Resized mosaic1073_.jpg to (224, 224)\n",
      "Resized mosaic1074_.jpg to (224, 224)\n",
      "Resized mosaic1075_.jpg to (224, 224)\n",
      "Resized mosaic1076_.jpg to (224, 224)\n",
      "Resized mosaic1077_.jpg to (224, 224)\n",
      "Resized mosaic1078_.jpg to (224, 224)\n",
      "Resized mosaic1079_.jpg to (224, 224)\n",
      "Resized mosaic107_.jpg to (224, 224)\n",
      "Resized mosaic1080_.jpg to (224, 224)\n",
      "Resized mosaic1081_.jpg to (224, 224)\n",
      "Resized mosaic1082_.jpg to (224, 224)\n",
      "Resized mosaic1083_.jpg to (224, 224)\n",
      "Resized mosaic1084_.jpg to (224, 224)\n",
      "Resized mosaic1085_.jpg to (224, 224)\n",
      "Resized mosaic1086_.jpg to (224, 224)\n",
      "Resized mosaic1087_.jpg to (224, 224)\n",
      "Resized mosaic1088_.jpg to (224, 224)\n",
      "Resized mosaic1089_.jpg to (224, 224)\n",
      "Resized mosaic108_.jpg to (224, 224)\n",
      "Resized mosaic1090_.jpg to (224, 224)\n",
      "Resized mosaic1091_.jpg to (224, 224)\n",
      "Resized mosaic1092_.jpg to (224, 224)\n",
      "Resized mosaic1093_.jpg to (224, 224)\n",
      "Resized mosaic1094_.jpg to (224, 224)\n",
      "Resized mosaic1095_.jpg to (224, 224)\n",
      "Resized mosaic1096_.jpg to (224, 224)\n",
      "Resized mosaic1097_.jpg to (224, 224)\n",
      "Resized mosaic1098_.jpg to (224, 224)\n",
      "Resized mosaic1099_.jpg to (224, 224)\n",
      "Resized mosaic109_.jpg to (224, 224)\n",
      "Resized mosaic10_.jpg to (224, 224)\n",
      "Resized mosaic1100_.jpg to (224, 224)\n",
      "Resized mosaic1101_.jpg to (224, 224)\n",
      "Resized mosaic1102_.jpg to (224, 224)\n",
      "Resized mosaic1103_.jpg to (224, 224)\n",
      "Resized mosaic1104_.jpg to (224, 224)\n",
      "Resized mosaic1105_.jpg to (224, 224)\n",
      "Resized mosaic1106_.jpg to (224, 224)\n",
      "Resized mosaic1107_.jpg to (224, 224)\n",
      "Resized mosaic1108_.jpg to (224, 224)\n",
      "Resized mosaic1109_.jpg to (224, 224)\n",
      "Resized mosaic110_.jpg to (224, 224)\n",
      "Resized mosaic1110_.jpg to (224, 224)\n",
      "Resized mosaic1111_.jpg to (224, 224)\n",
      "Resized mosaic1112_.jpg to (224, 224)\n",
      "Resized mosaic1113_.jpg to (224, 224)\n",
      "Resized mosaic1114_.jpg to (224, 224)\n",
      "Resized mosaic1115_.jpg to (224, 224)\n",
      "Resized mosaic1116_.jpg to (224, 224)\n",
      "Resized mosaic1117_.jpg to (224, 224)\n",
      "Resized mosaic1118_.jpg to (224, 224)\n",
      "Resized mosaic1119_.jpg to (224, 224)\n",
      "Resized mosaic111_.jpg to (224, 224)\n",
      "Resized mosaic1120_.jpg to (224, 224)\n",
      "Resized mosaic1121_.jpg to (224, 224)\n",
      "Resized mosaic1122_.jpg to (224, 224)\n",
      "Resized mosaic1123_.jpg to (224, 224)\n",
      "Resized mosaic1124_.jpg to (224, 224)\n",
      "Resized mosaic1125_.jpg to (224, 224)\n",
      "Resized mosaic1126_.jpg to (224, 224)\n",
      "Resized mosaic1127_.jpg to (224, 224)\n",
      "Resized mosaic1128_.jpg to (224, 224)\n",
      "Resized mosaic1129_.jpg to (224, 224)\n",
      "Resized mosaic112_.jpg to (224, 224)\n",
      "Resized mosaic1130_.jpg to (224, 224)\n",
      "Resized mosaic1131_.jpg to (224, 224)\n",
      "Resized mosaic1132_.jpg to (224, 224)\n",
      "Resized mosaic1133_.jpg to (224, 224)\n",
      "Resized mosaic1134_.jpg to (224, 224)\n",
      "Resized mosaic1135_.jpg to (224, 224)\n",
      "Resized mosaic1136_.jpg to (224, 224)\n",
      "Resized mosaic1137_.jpg to (224, 224)\n",
      "Resized mosaic1138_.jpg to (224, 224)\n",
      "Resized mosaic1139_.jpg to (224, 224)\n",
      "Resized mosaic113_.jpg to (224, 224)\n",
      "Resized mosaic1140_.jpg to (224, 224)\n",
      "Resized mosaic1141_.jpg to (224, 224)\n",
      "Resized mosaic1142_.jpg to (224, 224)\n",
      "Resized mosaic1143_.jpg to (224, 224)\n",
      "Resized mosaic1144_.jpg to (224, 224)\n",
      "Resized mosaic1145_.jpg to (224, 224)\n",
      "Resized mosaic1146_.jpg to (224, 224)\n",
      "Resized mosaic1147_.jpg to (224, 224)\n",
      "Resized mosaic1148_.jpg to (224, 224)\n",
      "Resized mosaic1149_.jpg to (224, 224)\n",
      "Resized mosaic114_.jpg to (224, 224)\n",
      "Resized mosaic1150_.jpg to (224, 224)\n",
      "Resized mosaic1151_.jpg to (224, 224)\n",
      "Resized mosaic1152_.jpg to (224, 224)\n",
      "Resized mosaic1153_.jpg to (224, 224)\n",
      "Resized mosaic1154_.jpg to (224, 224)\n",
      "Resized mosaic1155_.jpg to (224, 224)\n",
      "Resized mosaic1156_.jpg to (224, 224)\n",
      "Resized mosaic1157_.jpg to (224, 224)\n",
      "Resized mosaic1158_.jpg to (224, 224)\n",
      "Resized mosaic1159_.jpg to (224, 224)\n",
      "Resized mosaic115_.jpg to (224, 224)\n",
      "Resized mosaic1160_.jpg to (224, 224)\n",
      "Resized mosaic1161_.jpg to (224, 224)\n",
      "Resized mosaic1162_.jpg to (224, 224)\n",
      "Resized mosaic1163_.jpg to (224, 224)\n",
      "Resized mosaic1164_.jpg to (224, 224)\n",
      "Resized mosaic1165_.jpg to (224, 224)\n",
      "Resized mosaic1166_.jpg to (224, 224)\n",
      "Resized mosaic1167_.jpg to (224, 224)\n",
      "Resized mosaic1168_.jpg to (224, 224)\n",
      "Resized mosaic1169_.jpg to (224, 224)\n",
      "Resized mosaic116_.jpg to (224, 224)\n",
      "Resized mosaic1170_.jpg to (224, 224)\n",
      "Resized mosaic1171_.jpg to (224, 224)\n",
      "Resized mosaic1172_.jpg to (224, 224)\n",
      "Resized mosaic1173_.jpg to (224, 224)\n",
      "Resized mosaic1174_.jpg to (224, 224)\n",
      "Resized mosaic1175_.jpg to (224, 224)\n",
      "Resized mosaic1176_.jpg to (224, 224)\n",
      "Resized mosaic1177_.jpg to (224, 224)\n",
      "Resized mosaic1178_.jpg to (224, 224)\n",
      "Resized mosaic1179_.jpg to (224, 224)\n",
      "Resized mosaic117_.jpg to (224, 224)\n",
      "Resized mosaic1180_.jpg to (224, 224)\n",
      "Resized mosaic1181_.jpg to (224, 224)\n",
      "Resized mosaic1182_.jpg to (224, 224)\n",
      "Resized mosaic1183_.jpg to (224, 224)\n",
      "Resized mosaic1184_.jpg to (224, 224)\n",
      "Resized mosaic1185_.jpg to (224, 224)\n",
      "Resized mosaic1186_.jpg to (224, 224)\n",
      "Resized mosaic1187_.jpg to (224, 224)\n",
      "Resized mosaic1188_.jpg to (224, 224)\n",
      "Resized mosaic1189_.jpg to (224, 224)\n",
      "Resized mosaic118_.jpg to (224, 224)\n",
      "Resized mosaic1190_.jpg to (224, 224)\n",
      "Resized mosaic1191_.jpg to (224, 224)\n",
      "Resized mosaic1192_.jpg to (224, 224)\n",
      "Resized mosaic1193_.jpg to (224, 224)\n",
      "Resized mosaic1194_.jpg to (224, 224)\n",
      "Resized mosaic1195_.jpg to (224, 224)\n",
      "Resized mosaic1196_.jpg to (224, 224)\n",
      "Resized mosaic1197_.jpg to (224, 224)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resized mosaic1198_.jpg to (224, 224)\n",
      "Resized mosaic1199_.jpg to (224, 224)\n",
      "Resized mosaic119_.jpg to (224, 224)\n",
      "Resized mosaic11_.jpg to (224, 224)\n",
      "Resized mosaic1200_.jpg to (224, 224)\n",
      "Resized mosaic1201_.jpg to (224, 224)\n",
      "Resized mosaic1202_.jpg to (224, 224)\n",
      "Resized mosaic1203_.jpg to (224, 224)\n",
      "Resized mosaic1204_.jpg to (224, 224)\n",
      "Resized mosaic1205_.jpg to (224, 224)\n",
      "Resized mosaic1206_.jpg to (224, 224)\n",
      "Resized mosaic120_.jpg to (224, 224)\n",
      "Resized mosaic121_.jpg to (224, 224)\n",
      "Resized mosaic122_.jpg to (224, 224)\n",
      "Resized mosaic123_.jpg to (224, 224)\n",
      "Resized mosaic124_.jpg to (224, 224)\n",
      "Resized mosaic125_.jpg to (224, 224)\n",
      "Resized mosaic126_.jpg to (224, 224)\n",
      "Resized mosaic127_.jpg to (224, 224)\n",
      "Resized mosaic128_.jpg to (224, 224)\n",
      "Resized mosaic129_.jpg to (224, 224)\n",
      "Resized mosaic12_.jpg to (224, 224)\n",
      "Resized mosaic130_.jpg to (224, 224)\n",
      "Resized mosaic131_.jpg to (224, 224)\n",
      "Resized mosaic132_.jpg to (224, 224)\n",
      "Resized mosaic133_.jpg to (224, 224)\n",
      "Resized mosaic134_.jpg to (224, 224)\n",
      "Resized mosaic135_.jpg to (224, 224)\n",
      "Resized mosaic136_.jpg to (224, 224)\n",
      "Resized mosaic137_.jpg to (224, 224)\n",
      "Resized mosaic138_.jpg to (224, 224)\n",
      "Resized mosaic139_.jpg to (224, 224)\n",
      "Resized mosaic13_.jpg to (224, 224)\n",
      "Resized mosaic140_.jpg to (224, 224)\n",
      "Resized mosaic141_.jpg to (224, 224)\n",
      "Resized mosaic142_.jpg to (224, 224)\n",
      "Resized mosaic143_.jpg to (224, 224)\n",
      "Resized mosaic144_.jpg to (224, 224)\n",
      "Resized mosaic145_.jpg to (224, 224)\n",
      "Resized mosaic146_.jpg to (224, 224)\n",
      "Resized mosaic147_.jpg to (224, 224)\n",
      "Resized mosaic148_.jpg to (224, 224)\n",
      "Resized mosaic149_.jpg to (224, 224)\n",
      "Resized mosaic14_.jpg to (224, 224)\n",
      "Resized mosaic150_.jpg to (224, 224)\n",
      "Resized mosaic151_.jpg to (224, 224)\n",
      "Resized mosaic152_.jpg to (224, 224)\n",
      "Resized mosaic153_.jpg to (224, 224)\n",
      "Resized mosaic154_.jpg to (224, 224)\n",
      "Resized mosaic155_.jpg to (224, 224)\n",
      "Resized mosaic156_.jpg to (224, 224)\n",
      "Resized mosaic157_.jpg to (224, 224)\n",
      "Resized mosaic158_.jpg to (224, 224)\n",
      "Resized mosaic159_.jpg to (224, 224)\n",
      "Resized mosaic15_.jpg to (224, 224)\n",
      "Resized mosaic160_.jpg to (224, 224)\n",
      "Resized mosaic161_.jpg to (224, 224)\n",
      "Resized mosaic162_.jpg to (224, 224)\n",
      "Resized mosaic163_.jpg to (224, 224)\n",
      "Resized mosaic164_.jpg to (224, 224)\n",
      "Resized mosaic165_.jpg to (224, 224)\n",
      "Resized mosaic166_.jpg to (224, 224)\n",
      "Resized mosaic167_.jpg to (224, 224)\n",
      "Resized mosaic168_.jpg to (224, 224)\n",
      "Resized mosaic169_.jpg to (224, 224)\n",
      "Resized mosaic16_.jpg to (224, 224)\n",
      "Resized mosaic170_.jpg to (224, 224)\n",
      "Resized mosaic171_.jpg to (224, 224)\n",
      "Resized mosaic172_.jpg to (224, 224)\n",
      "Resized mosaic173_.jpg to (224, 224)\n",
      "Resized mosaic174_.jpg to (224, 224)\n",
      "Resized mosaic175_.jpg to (224, 224)\n",
      "Resized mosaic176_.jpg to (224, 224)\n",
      "Resized mosaic177_.jpg to (224, 224)\n",
      "Resized mosaic178_.jpg to (224, 224)\n",
      "Resized mosaic179_.jpg to (224, 224)\n",
      "Resized mosaic17_.jpg to (224, 224)\n",
      "Resized mosaic180_.jpg to (224, 224)\n",
      "Resized mosaic181_.jpg to (224, 224)\n",
      "Resized mosaic182_.jpg to (224, 224)\n",
      "Resized mosaic183_.jpg to (224, 224)\n",
      "Resized mosaic184_.jpg to (224, 224)\n",
      "Resized mosaic185_.jpg to (224, 224)\n",
      "Resized mosaic186_.jpg to (224, 224)\n",
      "Resized mosaic187_.jpg to (224, 224)\n",
      "Resized mosaic188_.jpg to (224, 224)\n",
      "Resized mosaic189_.jpg to (224, 224)\n",
      "Resized mosaic18_.jpg to (224, 224)\n",
      "Resized mosaic190_.jpg to (224, 224)\n",
      "Resized mosaic191_.jpg to (224, 224)\n",
      "Resized mosaic192_.jpg to (224, 224)\n",
      "Resized mosaic193_.jpg to (224, 224)\n",
      "Resized mosaic194_.jpg to (224, 224)\n",
      "Resized mosaic195_.jpg to (224, 224)\n",
      "Resized mosaic196_.jpg to (224, 224)\n",
      "Resized mosaic197_.jpg to (224, 224)\n",
      "Resized mosaic198_.jpg to (224, 224)\n",
      "Resized mosaic199_.jpg to (224, 224)\n",
      "Resized mosaic19_.jpg to (224, 224)\n",
      "Resized mosaic200_.jpg to (224, 224)\n",
      "Resized mosaic201_.jpg to (224, 224)\n",
      "Resized mosaic202_.jpg to (224, 224)\n",
      "Resized mosaic203_.jpg to (224, 224)\n",
      "Resized mosaic204_.jpg to (224, 224)\n",
      "Resized mosaic205_.jpg to (224, 224)\n",
      "Resized mosaic206_.jpg to (224, 224)\n",
      "Resized mosaic207_.jpg to (224, 224)\n",
      "Resized mosaic208_.jpg to (224, 224)\n",
      "Resized mosaic209_.jpg to (224, 224)\n",
      "Resized mosaic20_.jpg to (224, 224)\n",
      "Resized mosaic210_.jpg to (224, 224)\n",
      "Resized mosaic211_.jpg to (224, 224)\n",
      "Resized mosaic212_.jpg to (224, 224)\n",
      "Resized mosaic213_.jpg to (224, 224)\n",
      "Resized mosaic214_.jpg to (224, 224)\n",
      "Resized mosaic215_.jpg to (224, 224)\n",
      "Resized mosaic216_.jpg to (224, 224)\n",
      "Resized mosaic217_.jpg to (224, 224)\n",
      "Resized mosaic218_.jpg to (224, 224)\n",
      "Resized mosaic219_.jpg to (224, 224)\n",
      "Resized mosaic21_.jpg to (224, 224)\n",
      "Resized mosaic220_.jpg to (224, 224)\n",
      "Resized mosaic221_.jpg to (224, 224)\n",
      "Resized mosaic222_.jpg to (224, 224)\n",
      "Resized mosaic223_.jpg to (224, 224)\n",
      "Resized mosaic224_.jpg to (224, 224)\n",
      "Resized mosaic225_.jpg to (224, 224)\n",
      "Resized mosaic226_.jpg to (224, 224)\n",
      "Resized mosaic227_.jpg to (224, 224)\n",
      "Resized mosaic228_.jpg to (224, 224)\n",
      "Resized mosaic229_.jpg to (224, 224)\n",
      "Resized mosaic22_.jpg to (224, 224)\n",
      "Resized mosaic230_.jpg to (224, 224)\n",
      "Resized mosaic231_.jpg to (224, 224)\n",
      "Resized mosaic232_.jpg to (224, 224)\n",
      "Resized mosaic233_.jpg to (224, 224)\n",
      "Resized mosaic234_.jpg to (224, 224)\n",
      "Resized mosaic235_.jpg to (224, 224)\n",
      "Resized mosaic236_.jpg to (224, 224)\n",
      "Resized mosaic237_.jpg to (224, 224)\n",
      "Resized mosaic238_.jpg to (224, 224)\n",
      "Resized mosaic239_.jpg to (224, 224)\n",
      "Resized mosaic23_.jpg to (224, 224)\n",
      "Resized mosaic240_.jpg to (224, 224)\n",
      "Resized mosaic241_.jpg to (224, 224)\n",
      "Resized mosaic242_.jpg to (224, 224)\n",
      "Resized mosaic243_.jpg to (224, 224)\n",
      "Resized mosaic244_.jpg to (224, 224)\n",
      "Resized mosaic245_.jpg to (224, 224)\n",
      "Resized mosaic246_.jpg to (224, 224)\n",
      "Resized mosaic247_.jpg to (224, 224)\n",
      "Resized mosaic248_.jpg to (224, 224)\n",
      "Resized mosaic249_.jpg to (224, 224)\n",
      "Resized mosaic24_.jpg to (224, 224)\n",
      "Resized mosaic250_.jpg to (224, 224)\n",
      "Resized mosaic251_.jpg to (224, 224)\n",
      "Resized mosaic252_.jpg to (224, 224)\n",
      "Resized mosaic253_.jpg to (224, 224)\n",
      "Resized mosaic254_.jpg to (224, 224)\n",
      "Resized mosaic255_.jpg to (224, 224)\n",
      "Resized mosaic256_.jpg to (224, 224)\n",
      "Resized mosaic257_.jpg to (224, 224)\n",
      "Resized mosaic258_.jpg to (224, 224)\n",
      "Resized mosaic259_.jpg to (224, 224)\n",
      "Resized mosaic25_.jpg to (224, 224)\n",
      "Resized mosaic260_.jpg to (224, 224)\n",
      "Resized mosaic261_.jpg to (224, 224)\n",
      "Resized mosaic262_.jpg to (224, 224)\n",
      "Resized mosaic263_.jpg to (224, 224)\n",
      "Resized mosaic264_.jpg to (224, 224)\n",
      "Resized mosaic265_.jpg to (224, 224)\n",
      "Resized mosaic266_.jpg to (224, 224)\n",
      "Resized mosaic267_.jpg to (224, 224)\n",
      "Resized mosaic268_.jpg to (224, 224)\n",
      "Resized mosaic269_.jpg to (224, 224)\n",
      "Resized mosaic26_.jpg to (224, 224)\n",
      "Resized mosaic270_.jpg to (224, 224)\n",
      "Resized mosaic271_.jpg to (224, 224)\n",
      "Resized mosaic272_.jpg to (224, 224)\n",
      "Resized mosaic273_.jpg to (224, 224)\n",
      "Resized mosaic274_.jpg to (224, 224)\n",
      "Resized mosaic275_.jpg to (224, 224)\n",
      "Resized mosaic276_.jpg to (224, 224)\n",
      "Resized mosaic277_.jpg to (224, 224)\n",
      "Resized mosaic278_.jpg to (224, 224)\n",
      "Resized mosaic279_.jpg to (224, 224)\n",
      "Resized mosaic27_.jpg to (224, 224)\n",
      "Resized mosaic280_.jpg to (224, 224)\n",
      "Resized mosaic281_.jpg to (224, 224)\n",
      "Resized mosaic282_.jpg to (224, 224)\n",
      "Resized mosaic283_.jpg to (224, 224)\n",
      "Resized mosaic284_.jpg to (224, 224)\n",
      "Resized mosaic285_.jpg to (224, 224)\n",
      "Resized mosaic286_.jpg to (224, 224)\n",
      "Resized mosaic287_.jpg to (224, 224)\n",
      "Resized mosaic288_.jpg to (224, 224)\n",
      "Resized mosaic289_.jpg to (224, 224)\n",
      "Resized mosaic28_.jpg to (224, 224)\n",
      "Resized mosaic290_.jpg to (224, 224)\n",
      "Resized mosaic291_.jpg to (224, 224)\n",
      "Resized mosaic292_.jpg to (224, 224)\n",
      "Resized mosaic293_.jpg to (224, 224)\n",
      "Resized mosaic294_.jpg to (224, 224)\n",
      "Resized mosaic295_.jpg to (224, 224)\n",
      "Resized mosaic296_.jpg to (224, 224)\n",
      "Resized mosaic297_.jpg to (224, 224)\n",
      "Resized mosaic298_.jpg to (224, 224)\n",
      "Resized mosaic299_.jpg to (224, 224)\n",
      "Resized mosaic29_.jpg to (224, 224)\n",
      "Resized mosaic2_.jpg to (224, 224)\n",
      "Resized mosaic300_.jpg to (224, 224)\n",
      "Resized mosaic301_.jpg to (224, 224)\n",
      "Resized mosaic302_.jpg to (224, 224)\n",
      "Resized mosaic303_.jpg to (224, 224)\n",
      "Resized mosaic304_.jpg to (224, 224)\n",
      "Resized mosaic305_.jpg to (224, 224)\n",
      "Resized mosaic306_.jpg to (224, 224)\n",
      "Resized mosaic307_.jpg to (224, 224)\n",
      "Resized mosaic308_.jpg to (224, 224)\n",
      "Resized mosaic309_.jpg to (224, 224)\n",
      "Resized mosaic30_.jpg to (224, 224)\n",
      "Resized mosaic310_.jpg to (224, 224)\n",
      "Resized mosaic311_.jpg to (224, 224)\n",
      "Resized mosaic312_.jpg to (224, 224)\n",
      "Resized mosaic313_.jpg to (224, 224)\n",
      "Resized mosaic314_.jpg to (224, 224)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resized mosaic315_.jpg to (224, 224)\n",
      "Resized mosaic316_.jpg to (224, 224)\n",
      "Resized mosaic317_.jpg to (224, 224)\n",
      "Resized mosaic318_.jpg to (224, 224)\n",
      "Resized mosaic319_.jpg to (224, 224)\n",
      "Resized mosaic31_.jpg to (224, 224)\n",
      "Resized mosaic320_.jpg to (224, 224)\n",
      "Resized mosaic321_.jpg to (224, 224)\n",
      "Resized mosaic322_.jpg to (224, 224)\n",
      "Resized mosaic323_.jpg to (224, 224)\n",
      "Resized mosaic324_.jpg to (224, 224)\n",
      "Resized mosaic325_.jpg to (224, 224)\n",
      "Resized mosaic326_.jpg to (224, 224)\n",
      "Resized mosaic327_.jpg to (224, 224)\n",
      "Resized mosaic328_.jpg to (224, 224)\n",
      "Resized mosaic329_.jpg to (224, 224)\n",
      "Resized mosaic32_.jpg to (224, 224)\n",
      "Resized mosaic330_.jpg to (224, 224)\n",
      "Resized mosaic331_.jpg to (224, 224)\n",
      "Resized mosaic332_.jpg to (224, 224)\n",
      "Resized mosaic333_.jpg to (224, 224)\n",
      "Resized mosaic334_.jpg to (224, 224)\n",
      "Resized mosaic335_.jpg to (224, 224)\n",
      "Resized mosaic336_.jpg to (224, 224)\n",
      "Resized mosaic337_.jpg to (224, 224)\n",
      "Resized mosaic338_.jpg to (224, 224)\n",
      "Resized mosaic339_.jpg to (224, 224)\n",
      "Resized mosaic33_.jpg to (224, 224)\n",
      "Resized mosaic340_.jpg to (224, 224)\n",
      "Resized mosaic341_.jpg to (224, 224)\n",
      "Resized mosaic342_.jpg to (224, 224)\n",
      "Resized mosaic343_.jpg to (224, 224)\n",
      "Resized mosaic344_.jpg to (224, 224)\n",
      "Resized mosaic345_.jpg to (224, 224)\n",
      "Resized mosaic346_.jpg to (224, 224)\n",
      "Resized mosaic347_.jpg to (224, 224)\n",
      "Resized mosaic348_.jpg to (224, 224)\n",
      "Resized mosaic349_.jpg to (224, 224)\n",
      "Resized mosaic34_.jpg to (224, 224)\n",
      "Resized mosaic350_.jpg to (224, 224)\n",
      "Resized mosaic351_.jpg to (224, 224)\n",
      "Resized mosaic352_.jpg to (224, 224)\n",
      "Resized mosaic353_.jpg to (224, 224)\n",
      "Resized mosaic354_.jpg to (224, 224)\n",
      "Resized mosaic355_.jpg to (224, 224)\n",
      "Resized mosaic356_.jpg to (224, 224)\n",
      "Resized mosaic357_.jpg to (224, 224)\n",
      "Resized mosaic358_.jpg to (224, 224)\n",
      "Resized mosaic359_.jpg to (224, 224)\n",
      "Resized mosaic35_.jpg to (224, 224)\n",
      "Resized mosaic360_.jpg to (224, 224)\n",
      "Resized mosaic361_.jpg to (224, 224)\n",
      "Resized mosaic362_.jpg to (224, 224)\n",
      "Resized mosaic363_.jpg to (224, 224)\n",
      "Resized mosaic364_.jpg to (224, 224)\n",
      "Resized mosaic365_.jpg to (224, 224)\n",
      "Resized mosaic366_.jpg to (224, 224)\n",
      "Resized mosaic367_.jpg to (224, 224)\n",
      "Resized mosaic368_.jpg to (224, 224)\n",
      "Resized mosaic369_.jpg to (224, 224)\n",
      "Resized mosaic36_.jpg to (224, 224)\n",
      "Resized mosaic370_.jpg to (224, 224)\n",
      "Resized mosaic371_.jpg to (224, 224)\n",
      "Resized mosaic372_.jpg to (224, 224)\n",
      "Resized mosaic373_.jpg to (224, 224)\n",
      "Resized mosaic374_.jpg to (224, 224)\n",
      "Resized mosaic375_.jpg to (224, 224)\n",
      "Resized mosaic376_.jpg to (224, 224)\n",
      "Resized mosaic377_.jpg to (224, 224)\n",
      "Resized mosaic378_.jpg to (224, 224)\n",
      "Resized mosaic379_.jpg to (224, 224)\n",
      "Resized mosaic37_.jpg to (224, 224)\n",
      "Resized mosaic380_.jpg to (224, 224)\n",
      "Resized mosaic381_.jpg to (224, 224)\n",
      "Resized mosaic382_.jpg to (224, 224)\n",
      "Resized mosaic383_.jpg to (224, 224)\n",
      "Resized mosaic384_.jpg to (224, 224)\n",
      "Resized mosaic385_.jpg to (224, 224)\n",
      "Resized mosaic386_.jpg to (224, 224)\n",
      "Resized mosaic387_.jpg to (224, 224)\n",
      "Resized mosaic388_.jpg to (224, 224)\n",
      "Resized mosaic389_.jpg to (224, 224)\n",
      "Resized mosaic38_.jpg to (224, 224)\n",
      "Resized mosaic390_.jpg to (224, 224)\n",
      "Resized mosaic391_.jpg to (224, 224)\n",
      "Resized mosaic392_.jpg to (224, 224)\n",
      "Resized mosaic393_.jpg to (224, 224)\n",
      "Resized mosaic394_.jpg to (224, 224)\n",
      "Resized mosaic395_.jpg to (224, 224)\n",
      "Resized mosaic396_.jpg to (224, 224)\n",
      "Resized mosaic397_.jpg to (224, 224)\n",
      "Resized mosaic398_.jpg to (224, 224)\n",
      "Resized mosaic399_.jpg to (224, 224)\n",
      "Resized mosaic39_.jpg to (224, 224)\n",
      "Resized mosaic3_.jpg to (224, 224)\n",
      "Resized mosaic400_.jpg to (224, 224)\n",
      "Resized mosaic401_.jpg to (224, 224)\n",
      "Resized mosaic402_.jpg to (224, 224)\n",
      "Resized mosaic403_.jpg to (224, 224)\n",
      "Resized mosaic404_.jpg to (224, 224)\n",
      "Resized mosaic405_.jpg to (224, 224)\n",
      "Resized mosaic406_.jpg to (224, 224)\n",
      "Resized mosaic407_.jpg to (224, 224)\n",
      "Resized mosaic408_.jpg to (224, 224)\n",
      "Resized mosaic409_.jpg to (224, 224)\n",
      "Resized mosaic40_.jpg to (224, 224)\n",
      "Resized mosaic410_.jpg to (224, 224)\n",
      "Resized mosaic411_.jpg to (224, 224)\n",
      "Resized mosaic412_.jpg to (224, 224)\n",
      "Resized mosaic413_.jpg to (224, 224)\n",
      "Resized mosaic414_.jpg to (224, 224)\n",
      "Resized mosaic415_.jpg to (224, 224)\n",
      "Resized mosaic416_.jpg to (224, 224)\n",
      "Resized mosaic417_.jpg to (224, 224)\n",
      "Resized mosaic418_.jpg to (224, 224)\n",
      "Resized mosaic419_.jpg to (224, 224)\n",
      "Resized mosaic41_.jpg to (224, 224)\n",
      "Resized mosaic420_.jpg to (224, 224)\n",
      "Resized mosaic421_.jpg to (224, 224)\n",
      "Resized mosaic422_.jpg to (224, 224)\n",
      "Resized mosaic423_.jpg to (224, 224)\n",
      "Resized mosaic424_.jpg to (224, 224)\n",
      "Resized mosaic425_.jpg to (224, 224)\n",
      "Resized mosaic426_.jpg to (224, 224)\n",
      "Resized mosaic427_.jpg to (224, 224)\n",
      "Resized mosaic428_.jpg to (224, 224)\n",
      "Resized mosaic429_.jpg to (224, 224)\n",
      "Resized mosaic42_.jpg to (224, 224)\n",
      "Resized mosaic430_.jpg to (224, 224)\n",
      "Resized mosaic431_.jpg to (224, 224)\n",
      "Resized mosaic432_.jpg to (224, 224)\n",
      "Resized mosaic433_.jpg to (224, 224)\n",
      "Resized mosaic434_.jpg to (224, 224)\n",
      "Resized mosaic435_.jpg to (224, 224)\n",
      "Resized mosaic436_.jpg to (224, 224)\n",
      "Resized mosaic437_.jpg to (224, 224)\n",
      "Resized mosaic438_.jpg to (224, 224)\n",
      "Resized mosaic439_.jpg to (224, 224)\n",
      "Resized mosaic43_.jpg to (224, 224)\n",
      "Resized mosaic440_.jpg to (224, 224)\n",
      "Resized mosaic441_.jpg to (224, 224)\n",
      "Resized mosaic442_.jpg to (224, 224)\n",
      "Resized mosaic443_.jpg to (224, 224)\n",
      "Resized mosaic444_.jpg to (224, 224)\n",
      "Resized mosaic445_.jpg to (224, 224)\n",
      "Resized mosaic446_.jpg to (224, 224)\n",
      "Resized mosaic447_.jpg to (224, 224)\n",
      "Resized mosaic448_.jpg to (224, 224)\n",
      "Resized mosaic449_.jpg to (224, 224)\n",
      "Resized mosaic44_.jpg to (224, 224)\n",
      "Resized mosaic450_.jpg to (224, 224)\n",
      "Resized mosaic451_.jpg to (224, 224)\n",
      "Resized mosaic452_.jpg to (224, 224)\n",
      "Resized mosaic453_.jpg to (224, 224)\n",
      "Resized mosaic454_.jpg to (224, 224)\n",
      "Resized mosaic455_.jpg to (224, 224)\n",
      "Resized mosaic456_.jpg to (224, 224)\n",
      "Resized mosaic457_.jpg to (224, 224)\n",
      "Resized mosaic458_.jpg to (224, 224)\n",
      "Resized mosaic459_.jpg to (224, 224)\n",
      "Resized mosaic45_.jpg to (224, 224)\n",
      "Resized mosaic460_.jpg to (224, 224)\n",
      "Resized mosaic461_.jpg to (224, 224)\n",
      "Resized mosaic462_.jpg to (224, 224)\n",
      "Resized mosaic463_.jpg to (224, 224)\n",
      "Resized mosaic464_.jpg to (224, 224)\n",
      "Resized mosaic465_.jpg to (224, 224)\n",
      "Resized mosaic466_.jpg to (224, 224)\n",
      "Resized mosaic467_.jpg to (224, 224)\n",
      "Resized mosaic468_.jpg to (224, 224)\n",
      "Resized mosaic469_.jpg to (224, 224)\n",
      "Resized mosaic46_.jpg to (224, 224)\n",
      "Resized mosaic470_.jpg to (224, 224)\n",
      "Resized mosaic471_.jpg to (224, 224)\n",
      "Resized mosaic472_.jpg to (224, 224)\n",
      "Resized mosaic473_.jpg to (224, 224)\n",
      "Resized mosaic474_.jpg to (224, 224)\n",
      "Resized mosaic475_.jpg to (224, 224)\n",
      "Resized mosaic476_.jpg to (224, 224)\n",
      "Resized mosaic477_.jpg to (224, 224)\n",
      "Resized mosaic478_.jpg to (224, 224)\n",
      "Resized mosaic479_.jpg to (224, 224)\n",
      "Resized mosaic47_.jpg to (224, 224)\n",
      "Resized mosaic480_.jpg to (224, 224)\n",
      "Resized mosaic481_.jpg to (224, 224)\n",
      "Resized mosaic482_.jpg to (224, 224)\n",
      "Resized mosaic483_.jpg to (224, 224)\n",
      "Resized mosaic484_.jpg to (224, 224)\n",
      "Resized mosaic485_.jpg to (224, 224)\n",
      "Resized mosaic486_.jpg to (224, 224)\n",
      "Resized mosaic487_.jpg to (224, 224)\n",
      "Resized mosaic488_.jpg to (224, 224)\n",
      "Resized mosaic489_.jpg to (224, 224)\n",
      "Resized mosaic48_.jpg to (224, 224)\n",
      "Resized mosaic490_.jpg to (224, 224)\n",
      "Resized mosaic491_.jpg to (224, 224)\n",
      "Resized mosaic492_.jpg to (224, 224)\n",
      "Resized mosaic493_.jpg to (224, 224)\n",
      "Resized mosaic494_.jpg to (224, 224)\n",
      "Resized mosaic495_.jpg to (224, 224)\n",
      "Resized mosaic496_.jpg to (224, 224)\n",
      "Resized mosaic497_.jpg to (224, 224)\n",
      "Resized mosaic498_.jpg to (224, 224)\n",
      "Resized mosaic499_.jpg to (224, 224)\n",
      "Resized mosaic49_.jpg to (224, 224)\n",
      "Resized mosaic4_.jpg to (224, 224)\n",
      "Resized mosaic500_.jpg to (224, 224)\n",
      "Resized mosaic501_.jpg to (224, 224)\n",
      "Resized mosaic502_.jpg to (224, 224)\n",
      "Resized mosaic503_.jpg to (224, 224)\n",
      "Resized mosaic504_.jpg to (224, 224)\n",
      "Resized mosaic505_.jpg to (224, 224)\n",
      "Resized mosaic506_.jpg to (224, 224)\n",
      "Resized mosaic507_.jpg to (224, 224)\n",
      "Resized mosaic508_.jpg to (224, 224)\n",
      "Resized mosaic509_.jpg to (224, 224)\n",
      "Resized mosaic50_.jpg to (224, 224)\n",
      "Resized mosaic510_.jpg to (224, 224)\n",
      "Resized mosaic511_.jpg to (224, 224)\n",
      "Resized mosaic512_.jpg to (224, 224)\n",
      "Resized mosaic513_.jpg to (224, 224)\n",
      "Resized mosaic514_.jpg to (224, 224)\n",
      "Resized mosaic515_.jpg to (224, 224)\n",
      "Resized mosaic516_.jpg to (224, 224)\n",
      "Resized mosaic517_.jpg to (224, 224)\n",
      "Resized mosaic518_.jpg to (224, 224)\n",
      "Resized mosaic519_.jpg to (224, 224)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resized mosaic51_.jpg to (224, 224)\n",
      "Resized mosaic520_.jpg to (224, 224)\n",
      "Resized mosaic521_.jpg to (224, 224)\n",
      "Resized mosaic522_.jpg to (224, 224)\n",
      "Resized mosaic523_.jpg to (224, 224)\n",
      "Resized mosaic524_.jpg to (224, 224)\n",
      "Resized mosaic525_.jpg to (224, 224)\n",
      "Resized mosaic526_.jpg to (224, 224)\n",
      "Resized mosaic527_.jpg to (224, 224)\n",
      "Resized mosaic528_.jpg to (224, 224)\n",
      "Resized mosaic529_.jpg to (224, 224)\n",
      "Resized mosaic52_.jpg to (224, 224)\n",
      "Resized mosaic530_.jpg to (224, 224)\n",
      "Resized mosaic531_.jpg to (224, 224)\n",
      "Resized mosaic532_.jpg to (224, 224)\n",
      "Resized mosaic533_.jpg to (224, 224)\n",
      "Resized mosaic534_.jpg to (224, 224)\n",
      "Resized mosaic535_.jpg to (224, 224)\n",
      "Resized mosaic536_.jpg to (224, 224)\n",
      "Resized mosaic537_.jpg to (224, 224)\n",
      "Resized mosaic538_.jpg to (224, 224)\n",
      "Resized mosaic539_.jpg to (224, 224)\n",
      "Resized mosaic53_.jpg to (224, 224)\n",
      "Resized mosaic540_.jpg to (224, 224)\n",
      "Resized mosaic541_.jpg to (224, 224)\n",
      "Resized mosaic542_.jpg to (224, 224)\n",
      "Resized mosaic543_.jpg to (224, 224)\n",
      "Resized mosaic544_.jpg to (224, 224)\n",
      "Resized mosaic545_.jpg to (224, 224)\n",
      "Resized mosaic546_.jpg to (224, 224)\n",
      "Resized mosaic547_.jpg to (224, 224)\n",
      "Resized mosaic548_.jpg to (224, 224)\n",
      "Resized mosaic549_.jpg to (224, 224)\n",
      "Resized mosaic54_.jpg to (224, 224)\n",
      "Resized mosaic550_.jpg to (224, 224)\n",
      "Resized mosaic551_.jpg to (224, 224)\n",
      "Resized mosaic552_.jpg to (224, 224)\n",
      "Resized mosaic553_.jpg to (224, 224)\n",
      "Resized mosaic554_.jpg to (224, 224)\n",
      "Resized mosaic555_.jpg to (224, 224)\n",
      "Resized mosaic556_.jpg to (224, 224)\n",
      "Resized mosaic557_.jpg to (224, 224)\n",
      "Resized mosaic558_.jpg to (224, 224)\n",
      "Resized mosaic559_.jpg to (224, 224)\n",
      "Resized mosaic55_.jpg to (224, 224)\n",
      "Resized mosaic560_.jpg to (224, 224)\n",
      "Resized mosaic561_.jpg to (224, 224)\n",
      "Resized mosaic562_.jpg to (224, 224)\n",
      "Resized mosaic563_.jpg to (224, 224)\n",
      "Resized mosaic564_.jpg to (224, 224)\n",
      "Resized mosaic565_.jpg to (224, 224)\n",
      "Resized mosaic566_.jpg to (224, 224)\n",
      "Resized mosaic567_.jpg to (224, 224)\n",
      "Resized mosaic568_.jpg to (224, 224)\n",
      "Resized mosaic569_.jpg to (224, 224)\n",
      "Resized mosaic56_.jpg to (224, 224)\n",
      "Resized mosaic570_.jpg to (224, 224)\n",
      "Resized mosaic571_.jpg to (224, 224)\n",
      "Resized mosaic572_.jpg to (224, 224)\n",
      "Resized mosaic573_.jpg to (224, 224)\n",
      "Resized mosaic574_.jpg to (224, 224)\n",
      "Resized mosaic575_.jpg to (224, 224)\n",
      "Resized mosaic576_.jpg to (224, 224)\n",
      "Resized mosaic577_.jpg to (224, 224)\n",
      "Resized mosaic578_.jpg to (224, 224)\n",
      "Resized mosaic579_.jpg to (224, 224)\n",
      "Resized mosaic57_.jpg to (224, 224)\n",
      "Resized mosaic580_.jpg to (224, 224)\n",
      "Resized mosaic581_.jpg to (224, 224)\n",
      "Resized mosaic582_.jpg to (224, 224)\n",
      "Resized mosaic583_.jpg to (224, 224)\n",
      "Resized mosaic584_.jpg to (224, 224)\n",
      "Resized mosaic585_.jpg to (224, 224)\n",
      "Resized mosaic586_.jpg to (224, 224)\n",
      "Resized mosaic587_.jpg to (224, 224)\n",
      "Resized mosaic588_.jpg to (224, 224)\n",
      "Resized mosaic589_.jpg to (224, 224)\n",
      "Resized mosaic58_.jpg to (224, 224)\n",
      "Resized mosaic590_.jpg to (224, 224)\n",
      "Resized mosaic591_.jpg to (224, 224)\n",
      "Resized mosaic592_.jpg to (224, 224)\n",
      "Resized mosaic593_.jpg to (224, 224)\n",
      "Resized mosaic594_.jpg to (224, 224)\n",
      "Resized mosaic595_.jpg to (224, 224)\n",
      "Resized mosaic596_.jpg to (224, 224)\n",
      "Resized mosaic597_.jpg to (224, 224)\n",
      "Resized mosaic598_.jpg to (224, 224)\n",
      "Resized mosaic599_.jpg to (224, 224)\n",
      "Resized mosaic59_.jpg to (224, 224)\n",
      "Resized mosaic5_.jpg to (224, 224)\n",
      "Resized mosaic600_.jpg to (224, 224)\n",
      "Resized mosaic601_.jpg to (224, 224)\n",
      "Resized mosaic602_.jpg to (224, 224)\n",
      "Resized mosaic603_.jpg to (224, 224)\n",
      "Resized mosaic604_.jpg to (224, 224)\n",
      "Resized mosaic605_.jpg to (224, 224)\n",
      "Resized mosaic606_.jpg to (224, 224)\n",
      "Resized mosaic607_.jpg to (224, 224)\n",
      "Resized mosaic608_.jpg to (224, 224)\n",
      "Resized mosaic609_.jpg to (224, 224)\n",
      "Resized mosaic60_.jpg to (224, 224)\n",
      "Resized mosaic610_.jpg to (224, 224)\n",
      "Resized mosaic611_.jpg to (224, 224)\n",
      "Resized mosaic612_.jpg to (224, 224)\n",
      "Resized mosaic613_.jpg to (224, 224)\n",
      "Resized mosaic614_.jpg to (224, 224)\n",
      "Resized mosaic615_.jpg to (224, 224)\n",
      "Resized mosaic616_.jpg to (224, 224)\n",
      "Resized mosaic617_.jpg to (224, 224)\n",
      "Resized mosaic618_.jpg to (224, 224)\n",
      "Resized mosaic619_.jpg to (224, 224)\n",
      "Resized mosaic61_.jpg to (224, 224)\n",
      "Resized mosaic620_.jpg to (224, 224)\n",
      "Resized mosaic621_.jpg to (224, 224)\n",
      "Resized mosaic622_.jpg to (224, 224)\n",
      "Resized mosaic623_.jpg to (224, 224)\n",
      "Resized mosaic624_.jpg to (224, 224)\n",
      "Resized mosaic625_.jpg to (224, 224)\n",
      "Resized mosaic62_.jpg to (224, 224)\n",
      "Resized mosaic632_.jpg to (224, 224)\n",
      "Resized mosaic633_.jpg to (224, 224)\n",
      "Resized mosaic634_.jpg to (224, 224)\n",
      "Resized mosaic635_.jpg to (224, 224)\n",
      "Resized mosaic636_.jpg to (224, 224)\n",
      "Resized mosaic637_.jpg to (224, 224)\n",
      "Resized mosaic638_.jpg to (224, 224)\n",
      "Resized mosaic639_.jpg to (224, 224)\n",
      "Resized mosaic63_.jpg to (224, 224)\n",
      "Resized mosaic640_.jpg to (224, 224)\n",
      "Resized mosaic641_.jpg to (224, 224)\n",
      "Resized mosaic642_.jpg to (224, 224)\n",
      "Resized mosaic643_.jpg to (224, 224)\n",
      "Resized mosaic644_.jpg to (224, 224)\n",
      "Resized mosaic645_.jpg to (224, 224)\n",
      "Resized mosaic646_.jpg to (224, 224)\n",
      "Resized mosaic647_.jpg to (224, 224)\n",
      "Resized mosaic648_.jpg to (224, 224)\n",
      "Resized mosaic649_.jpg to (224, 224)\n",
      "Resized mosaic64_.jpg to (224, 224)\n",
      "Resized mosaic650_.jpg to (224, 224)\n",
      "Resized mosaic651_.jpg to (224, 224)\n",
      "Resized mosaic652_.jpg to (224, 224)\n",
      "Resized mosaic653_.jpg to (224, 224)\n",
      "Resized mosaic654_.jpg to (224, 224)\n",
      "Resized mosaic655_.jpg to (224, 224)\n",
      "Resized mosaic656_.jpg to (224, 224)\n",
      "Resized mosaic657_.jpg to (224, 224)\n",
      "Resized mosaic658_.jpg to (224, 224)\n",
      "Resized mosaic659_.jpg to (224, 224)\n",
      "Resized mosaic65_.jpg to (224, 224)\n",
      "Resized mosaic660_.jpg to (224, 224)\n",
      "Resized mosaic661_.jpg to (224, 224)\n",
      "Resized mosaic662_.jpg to (224, 224)\n",
      "Resized mosaic663_.jpg to (224, 224)\n",
      "Resized mosaic664_.jpg to (224, 224)\n",
      "Resized mosaic665_.jpg to (224, 224)\n",
      "Resized mosaic666_.jpg to (224, 224)\n",
      "Resized mosaic667_.jpg to (224, 224)\n",
      "Resized mosaic668_.jpg to (224, 224)\n",
      "Resized mosaic669_.jpg to (224, 224)\n",
      "Resized mosaic66_.jpg to (224, 224)\n",
      "Resized mosaic670_.jpg to (224, 224)\n",
      "Resized mosaic671_.jpg to (224, 224)\n",
      "Resized mosaic672_.jpg to (224, 224)\n",
      "Resized mosaic673_.jpg to (224, 224)\n",
      "Resized mosaic674_.jpg to (224, 224)\n",
      "Resized mosaic675_.jpg to (224, 224)\n",
      "Resized mosaic676_.jpg to (224, 224)\n",
      "Resized mosaic677_.jpg to (224, 224)\n",
      "Resized mosaic678_.jpg to (224, 224)\n",
      "Resized mosaic679_.jpg to (224, 224)\n",
      "Resized mosaic67_.jpg to (224, 224)\n",
      "Resized mosaic680_.jpg to (224, 224)\n",
      "Resized mosaic681_.jpg to (224, 224)\n",
      "Resized mosaic682_.jpg to (224, 224)\n",
      "Resized mosaic683_.jpg to (224, 224)\n",
      "Resized mosaic684_.jpg to (224, 224)\n",
      "Resized mosaic685_.jpg to (224, 224)\n",
      "Resized mosaic686_.jpg to (224, 224)\n",
      "Resized mosaic687_.jpg to (224, 224)\n",
      "Resized mosaic688_.jpg to (224, 224)\n",
      "Resized mosaic689_.jpg to (224, 224)\n",
      "Resized mosaic68_.jpg to (224, 224)\n",
      "Resized mosaic690_.jpg to (224, 224)\n",
      "Resized mosaic691_.jpg to (224, 224)\n",
      "Resized mosaic692_.jpg to (224, 224)\n",
      "Resized mosaic693_.jpg to (224, 224)\n",
      "Resized mosaic694_.jpg to (224, 224)\n",
      "Resized mosaic695_.jpg to (224, 224)\n",
      "Resized mosaic696_.jpg to (224, 224)\n",
      "Resized mosaic697_.jpg to (224, 224)\n",
      "Resized mosaic698_.jpg to (224, 224)\n",
      "Resized mosaic699_.jpg to (224, 224)\n",
      "Resized mosaic69_.jpg to (224, 224)\n",
      "Resized mosaic6_.jpg to (224, 224)\n",
      "Resized mosaic700_.jpg to (224, 224)\n",
      "Resized mosaic701_.jpg to (224, 224)\n",
      "Resized mosaic702_.jpg to (224, 224)\n",
      "Resized mosaic703_.jpg to (224, 224)\n",
      "Resized mosaic704_.jpg to (224, 224)\n",
      "Resized mosaic705_.jpg to (224, 224)\n",
      "Resized mosaic706_.jpg to (224, 224)\n",
      "Resized mosaic707_.jpg to (224, 224)\n",
      "Resized mosaic708_.jpg to (224, 224)\n",
      "Resized mosaic709_.jpg to (224, 224)\n",
      "Resized mosaic70_.jpg to (224, 224)\n",
      "Resized mosaic710_.jpg to (224, 224)\n",
      "Resized mosaic711_.jpg to (224, 224)\n",
      "Resized mosaic712_.jpg to (224, 224)\n",
      "Resized mosaic713_.jpg to (224, 224)\n",
      "Resized mosaic714_.jpg to (224, 224)\n",
      "Resized mosaic715_.jpg to (224, 224)\n",
      "Resized mosaic716_.jpg to (224, 224)\n",
      "Resized mosaic717_.jpg to (224, 224)\n",
      "Resized mosaic718_.jpg to (224, 224)\n",
      "Resized mosaic719_.jpg to (224, 224)\n",
      "Resized mosaic71_.jpg to (224, 224)\n",
      "Resized mosaic720_.jpg to (224, 224)\n",
      "Resized mosaic721_.jpg to (224, 224)\n",
      "Resized mosaic722_.jpg to (224, 224)\n",
      "Resized mosaic723_.jpg to (224, 224)\n",
      "Resized mosaic724_.jpg to (224, 224)\n",
      "Resized mosaic725_.jpg to (224, 224)\n",
      "Resized mosaic726_.jpg to (224, 224)\n",
      "Resized mosaic727_.jpg to (224, 224)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resized mosaic728_.jpg to (224, 224)\n",
      "Resized mosaic729_.jpg to (224, 224)\n",
      "Resized mosaic72_.jpg to (224, 224)\n",
      "Resized mosaic730_.jpg to (224, 224)\n",
      "Resized mosaic731_.jpg to (224, 224)\n",
      "Resized mosaic732_.jpg to (224, 224)\n",
      "Resized mosaic733_.jpg to (224, 224)\n",
      "Resized mosaic734_.jpg to (224, 224)\n",
      "Resized mosaic735_.jpg to (224, 224)\n",
      "Resized mosaic736_.jpg to (224, 224)\n",
      "Resized mosaic737_.jpg to (224, 224)\n",
      "Resized mosaic738_.jpg to (224, 224)\n",
      "Resized mosaic739_.jpg to (224, 224)\n",
      "Resized mosaic73_.jpg to (224, 224)\n",
      "Resized mosaic740_.jpg to (224, 224)\n",
      "Resized mosaic741_.jpg to (224, 224)\n",
      "Resized mosaic742_.jpg to (224, 224)\n",
      "Resized mosaic743_.jpg to (224, 224)\n",
      "Resized mosaic744_.jpg to (224, 224)\n",
      "Resized mosaic745_.jpg to (224, 224)\n",
      "Resized mosaic746_.jpg to (224, 224)\n",
      "Resized mosaic747_.jpg to (224, 224)\n",
      "Resized mosaic748_.jpg to (224, 224)\n",
      "Resized mosaic749_.jpg to (224, 224)\n",
      "Resized mosaic74_.jpg to (224, 224)\n",
      "Resized mosaic750_.jpg to (224, 224)\n",
      "Resized mosaic751_.jpg to (224, 224)\n",
      "Resized mosaic752_.jpg to (224, 224)\n",
      "Resized mosaic753_.jpg to (224, 224)\n",
      "Resized mosaic754_.jpg to (224, 224)\n",
      "Resized mosaic755_.jpg to (224, 224)\n",
      "Resized mosaic756_.jpg to (224, 224)\n",
      "Resized mosaic757_.jpg to (224, 224)\n",
      "Resized mosaic758_.jpg to (224, 224)\n",
      "Resized mosaic759_.jpg to (224, 224)\n",
      "Resized mosaic75_.jpg to (224, 224)\n",
      "Resized mosaic760_.jpg to (224, 224)\n",
      "Resized mosaic761_.jpg to (224, 224)\n",
      "Resized mosaic762_.jpg to (224, 224)\n",
      "Resized mosaic763_.jpg to (224, 224)\n",
      "Resized mosaic764_.jpg to (224, 224)\n",
      "Resized mosaic765_.jpg to (224, 224)\n",
      "Resized mosaic766_.jpg to (224, 224)\n",
      "Resized mosaic767_.jpg to (224, 224)\n",
      "Resized mosaic768_.jpg to (224, 224)\n",
      "Resized mosaic769_.jpg to (224, 224)\n",
      "Resized mosaic76_.jpg to (224, 224)\n",
      "Resized mosaic770_.jpg to (224, 224)\n",
      "Resized mosaic771_.jpg to (224, 224)\n",
      "Resized mosaic772_.jpg to (224, 224)\n",
      "Resized mosaic773_.jpg to (224, 224)\n",
      "Resized mosaic774_.jpg to (224, 224)\n",
      "Resized mosaic775_.jpg to (224, 224)\n",
      "Resized mosaic776_.jpg to (224, 224)\n",
      "Resized mosaic777_.jpg to (224, 224)\n",
      "Resized mosaic778_.jpg to (224, 224)\n",
      "Resized mosaic779_.jpg to (224, 224)\n",
      "Resized mosaic77_.jpg to (224, 224)\n",
      "Resized mosaic780_.jpg to (224, 224)\n",
      "Resized mosaic781_.jpg to (224, 224)\n",
      "Resized mosaic782_.jpg to (224, 224)\n",
      "Resized mosaic783_.jpg to (224, 224)\n",
      "Resized mosaic784_.jpg to (224, 224)\n",
      "Resized mosaic785_.jpg to (224, 224)\n",
      "Resized mosaic786_.jpg to (224, 224)\n",
      "Resized mosaic787_.jpg to (224, 224)\n",
      "Resized mosaic788_.jpg to (224, 224)\n",
      "Resized mosaic789_.jpg to (224, 224)\n",
      "Resized mosaic78_.jpg to (224, 224)\n",
      "Resized mosaic790_.jpg to (224, 224)\n",
      "Resized mosaic791_.jpg to (224, 224)\n",
      "Resized mosaic792_.jpg to (224, 224)\n",
      "Resized mosaic793_.jpg to (224, 224)\n",
      "Resized mosaic794_.jpg to (224, 224)\n",
      "Resized mosaic795_.jpg to (224, 224)\n",
      "Resized mosaic796_.jpg to (224, 224)\n",
      "Resized mosaic797_.jpg to (224, 224)\n",
      "Resized mosaic798_.jpg to (224, 224)\n",
      "Resized mosaic799_.jpg to (224, 224)\n",
      "Resized mosaic79_.jpg to (224, 224)\n",
      "Resized mosaic7_.jpg to (224, 224)\n",
      "Resized mosaic800_.jpg to (224, 224)\n",
      "Resized mosaic801_.jpg to (224, 224)\n",
      "Resized mosaic802_.jpg to (224, 224)\n",
      "Resized mosaic803_.jpg to (224, 224)\n",
      "Resized mosaic804_.jpg to (224, 224)\n",
      "Resized mosaic805_.jpg to (224, 224)\n",
      "Resized mosaic806_.jpg to (224, 224)\n",
      "Resized mosaic807_.jpg to (224, 224)\n",
      "Resized mosaic808_.jpg to (224, 224)\n",
      "Resized mosaic809_.jpg to (224, 224)\n",
      "Resized mosaic80_.jpg to (224, 224)\n",
      "Resized mosaic810_.jpg to (224, 224)\n",
      "Resized mosaic811_.jpg to (224, 224)\n",
      "Resized mosaic812_.jpg to (224, 224)\n",
      "Resized mosaic813_.jpg to (224, 224)\n",
      "Resized mosaic814_.jpg to (224, 224)\n",
      "Resized mosaic815_.jpg to (224, 224)\n",
      "Resized mosaic816_.jpg to (224, 224)\n",
      "Resized mosaic817_.jpg to (224, 224)\n",
      "Resized mosaic818_.jpg to (224, 224)\n",
      "Resized mosaic819_.jpg to (224, 224)\n",
      "Resized mosaic81_.jpg to (224, 224)\n",
      "Resized mosaic820_.jpg to (224, 224)\n",
      "Resized mosaic821_.jpg to (224, 224)\n",
      "Resized mosaic822_.jpg to (224, 224)\n",
      "Resized mosaic823_.jpg to (224, 224)\n",
      "Resized mosaic824_.jpg to (224, 224)\n",
      "Resized mosaic825_.jpg to (224, 224)\n",
      "Resized mosaic826_.jpg to (224, 224)\n",
      "Resized mosaic827_.jpg to (224, 224)\n",
      "Resized mosaic828_.jpg to (224, 224)\n",
      "Resized mosaic829_.jpg to (224, 224)\n",
      "Resized mosaic82_.jpg to (224, 224)\n",
      "Resized mosaic830_.jpg to (224, 224)\n",
      "Resized mosaic831_.jpg to (224, 224)\n",
      "Resized mosaic832_.jpg to (224, 224)\n",
      "Resized mosaic833_.jpg to (224, 224)\n",
      "Resized mosaic834_.jpg to (224, 224)\n",
      "Resized mosaic835_.jpg to (224, 224)\n",
      "Resized mosaic836_.jpg to (224, 224)\n",
      "Resized mosaic837_.jpg to (224, 224)\n",
      "Resized mosaic838_.jpg to (224, 224)\n",
      "Resized mosaic839_.jpg to (224, 224)\n",
      "Resized mosaic83_.jpg to (224, 224)\n",
      "Resized mosaic840_.jpg to (224, 224)\n",
      "Resized mosaic841_.jpg to (224, 224)\n",
      "Resized mosaic842_.jpg to (224, 224)\n",
      "Resized mosaic843_.jpg to (224, 224)\n",
      "Resized mosaic844_.jpg to (224, 224)\n",
      "Resized mosaic845_.jpg to (224, 224)\n",
      "Resized mosaic846_.jpg to (224, 224)\n",
      "Resized mosaic847_.jpg to (224, 224)\n",
      "Resized mosaic848_.jpg to (224, 224)\n",
      "Resized mosaic849_.jpg to (224, 224)\n",
      "Resized mosaic84_.jpg to (224, 224)\n",
      "Resized mosaic850_.jpg to (224, 224)\n",
      "Resized mosaic851_.jpg to (224, 224)\n",
      "Resized mosaic852_.jpg to (224, 224)\n",
      "Resized mosaic853_.jpg to (224, 224)\n",
      "Resized mosaic854_.jpg to (224, 224)\n",
      "Resized mosaic855_.jpg to (224, 224)\n",
      "Resized mosaic856_.jpg to (224, 224)\n",
      "Resized mosaic857_.jpg to (224, 224)\n",
      "Resized mosaic858_.jpg to (224, 224)\n",
      "Resized mosaic859_.jpg to (224, 224)\n",
      "Resized mosaic85_.jpg to (224, 224)\n",
      "Resized mosaic860_.jpg to (224, 224)\n",
      "Resized mosaic861_.jpg to (224, 224)\n",
      "Resized mosaic862_.jpg to (224, 224)\n",
      "Resized mosaic863_.jpg to (224, 224)\n",
      "Resized mosaic864_.jpg to (224, 224)\n",
      "Resized mosaic865_.jpg to (224, 224)\n",
      "Resized mosaic866_.jpg to (224, 224)\n",
      "Resized mosaic867_.jpg to (224, 224)\n",
      "Resized mosaic868_.jpg to (224, 224)\n",
      "Resized mosaic869_.jpg to (224, 224)\n",
      "Resized mosaic86_.jpg to (224, 224)\n",
      "Resized mosaic870_.jpg to (224, 224)\n",
      "Resized mosaic871_.jpg to (224, 224)\n",
      "Resized mosaic872_.jpg to (224, 224)\n",
      "Resized mosaic873_.jpg to (224, 224)\n",
      "Resized mosaic874_.jpg to (224, 224)\n",
      "Resized mosaic875_.jpg to (224, 224)\n",
      "Resized mosaic876_.jpg to (224, 224)\n",
      "Resized mosaic877_.jpg to (224, 224)\n",
      "Resized mosaic878_.jpg to (224, 224)\n",
      "Resized mosaic879_.jpg to (224, 224)\n",
      "Resized mosaic87_.jpg to (224, 224)\n",
      "Resized mosaic880_.jpg to (224, 224)\n",
      "Resized mosaic881_.jpg to (224, 224)\n",
      "Resized mosaic882_.jpg to (224, 224)\n",
      "Resized mosaic883_.jpg to (224, 224)\n",
      "Resized mosaic884_.jpg to (224, 224)\n",
      "Resized mosaic885_.jpg to (224, 224)\n",
      "Resized mosaic886_.jpg to (224, 224)\n",
      "Resized mosaic887_.jpg to (224, 224)\n",
      "Resized mosaic888_.jpg to (224, 224)\n",
      "Resized mosaic889_.jpg to (224, 224)\n",
      "Resized mosaic88_.jpg to (224, 224)\n",
      "Resized mosaic890_.jpg to (224, 224)\n",
      "Resized mosaic891_.jpg to (224, 224)\n",
      "Resized mosaic892_.jpg to (224, 224)\n",
      "Resized mosaic893_.jpg to (224, 224)\n",
      "Resized mosaic894_.jpg to (224, 224)\n",
      "Resized mosaic895_.jpg to (224, 224)\n",
      "Resized mosaic896_.jpg to (224, 224)\n",
      "Resized mosaic897_.jpg to (224, 224)\n",
      "Resized mosaic898_.jpg to (224, 224)\n",
      "Resized mosaic899_.jpg to (224, 224)\n",
      "Resized mosaic89_.jpg to (224, 224)\n",
      "Resized mosaic8_.jpg to (224, 224)\n",
      "Resized mosaic900_.jpg to (224, 224)\n",
      "Resized mosaic901_.jpg to (224, 224)\n",
      "Resized mosaic902_.jpg to (224, 224)\n",
      "Resized mosaic903_.jpg to (224, 224)\n",
      "Resized mosaic904_.jpg to (224, 224)\n",
      "Resized mosaic905_.jpg to (224, 224)\n",
      "Resized mosaic906_.jpg to (224, 224)\n",
      "Resized mosaic907_.jpg to (224, 224)\n",
      "Resized mosaic908_.jpg to (224, 224)\n",
      "Resized mosaic909_.jpg to (224, 224)\n",
      "Resized mosaic90_.jpg to (224, 224)\n",
      "Resized mosaic910_.jpg to (224, 224)\n",
      "Resized mosaic911_.jpg to (224, 224)\n",
      "Resized mosaic912_.jpg to (224, 224)\n",
      "Resized mosaic913_.jpg to (224, 224)\n",
      "Resized mosaic914_.jpg to (224, 224)\n",
      "Resized mosaic915_.jpg to (224, 224)\n",
      "Resized mosaic916_.jpg to (224, 224)\n",
      "Resized mosaic917_.jpg to (224, 224)\n",
      "Resized mosaic918_.jpg to (224, 224)\n",
      "Resized mosaic919_.jpg to (224, 224)\n",
      "Resized mosaic91_.jpg to (224, 224)\n",
      "Resized mosaic920_.jpg to (224, 224)\n",
      "Resized mosaic921_.jpg to (224, 224)\n",
      "Resized mosaic922_.jpg to (224, 224)\n",
      "Resized mosaic923_.jpg to (224, 224)\n",
      "Resized mosaic924_.jpg to (224, 224)\n",
      "Resized mosaic925_.jpg to (224, 224)\n",
      "Resized mosaic926_.jpg to (224, 224)\n",
      "Resized mosaic927_.jpg to (224, 224)\n",
      "Resized mosaic928_.jpg to (224, 224)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resized mosaic929_.jpg to (224, 224)\n",
      "Resized mosaic92_.jpg to (224, 224)\n",
      "Resized mosaic930_.jpg to (224, 224)\n",
      "Resized mosaic931_.jpg to (224, 224)\n",
      "Resized mosaic932_.jpg to (224, 224)\n",
      "Resized mosaic933_.jpg to (224, 224)\n",
      "Resized mosaic934_.jpg to (224, 224)\n",
      "Resized mosaic935_.jpg to (224, 224)\n",
      "Resized mosaic936_.jpg to (224, 224)\n",
      "Resized mosaic937_.jpg to (224, 224)\n",
      "Resized mosaic938_.jpg to (224, 224)\n",
      "Resized mosaic939_.jpg to (224, 224)\n",
      "Resized mosaic93_.jpg to (224, 224)\n",
      "Resized mosaic940_.jpg to (224, 224)\n",
      "Resized mosaic941_.jpg to (224, 224)\n",
      "Resized mosaic942_.jpg to (224, 224)\n",
      "Resized mosaic943_.jpg to (224, 224)\n",
      "Resized mosaic944_.jpg to (224, 224)\n",
      "Resized mosaic945_.jpg to (224, 224)\n",
      "Resized mosaic946_.jpg to (224, 224)\n",
      "Resized mosaic947_.jpg to (224, 224)\n",
      "Resized mosaic948_.jpg to (224, 224)\n",
      "Resized mosaic949_.jpg to (224, 224)\n",
      "Resized mosaic94_.jpg to (224, 224)\n",
      "Resized mosaic950_.jpg to (224, 224)\n",
      "Resized mosaic951_.jpg to (224, 224)\n",
      "Resized mosaic952_.jpg to (224, 224)\n",
      "Resized mosaic953_.jpg to (224, 224)\n",
      "Resized mosaic954_.jpg to (224, 224)\n",
      "Resized mosaic955_.jpg to (224, 224)\n",
      "Resized mosaic956_.jpg to (224, 224)\n",
      "Resized mosaic957_.jpg to (224, 224)\n",
      "Resized mosaic958_.jpg to (224, 224)\n",
      "Resized mosaic959_.jpg to (224, 224)\n",
      "Resized mosaic95_.jpg to (224, 224)\n",
      "Resized mosaic960_.jpg to (224, 224)\n",
      "Resized mosaic961_.jpg to (224, 224)\n",
      "Resized mosaic962_.jpg to (224, 224)\n",
      "Resized mosaic963_.jpg to (224, 224)\n",
      "Resized mosaic964_.jpg to (224, 224)\n",
      "Resized mosaic965_.jpg to (224, 224)\n",
      "Resized mosaic966_.jpg to (224, 224)\n",
      "Resized mosaic967_.jpg to (224, 224)\n",
      "Resized mosaic968_.jpg to (224, 224)\n",
      "Resized mosaic969_.jpg to (224, 224)\n",
      "Resized mosaic96_.jpg to (224, 224)\n",
      "Resized mosaic970_.jpg to (224, 224)\n",
      "Resized mosaic971_.jpg to (224, 224)\n",
      "Resized mosaic972_.jpg to (224, 224)\n",
      "Resized mosaic973_.jpg to (224, 224)\n",
      "Resized mosaic974_.jpg to (224, 224)\n",
      "Resized mosaic975_.jpg to (224, 224)\n",
      "Resized mosaic976_.jpg to (224, 224)\n",
      "Resized mosaic977_.jpg to (224, 224)\n",
      "Resized mosaic978_.jpg to (224, 224)\n",
      "Resized mosaic979_.jpg to (224, 224)\n",
      "Resized mosaic97_.jpg to (224, 224)\n",
      "Resized mosaic980_.jpg to (224, 224)\n",
      "Resized mosaic981_.jpg to (224, 224)\n",
      "Resized mosaic982_.jpg to (224, 224)\n",
      "Resized mosaic983_.jpg to (224, 224)\n",
      "Resized mosaic984_.jpg to (224, 224)\n",
      "Resized mosaic985_.jpg to (224, 224)\n",
      "Resized mosaic986_.jpg to (224, 224)\n",
      "Resized mosaic987_.jpg to (224, 224)\n",
      "Resized mosaic988_.jpg to (224, 224)\n",
      "Resized mosaic989_.jpg to (224, 224)\n",
      "Resized mosaic98_.jpg to (224, 224)\n",
      "Resized mosaic990_.jpg to (224, 224)\n",
      "Resized mosaic991_.jpg to (224, 224)\n",
      "Resized mosaic992_.jpg to (224, 224)\n",
      "Resized mosaic993_.jpg to (224, 224)\n",
      "Resized mosaic994_.jpg to (224, 224)\n",
      "Resized mosaic995_.jpg to (224, 224)\n",
      "Resized mosaic996_.jpg to (224, 224)\n",
      "Resized mosaic997_.jpg to (224, 224)\n",
      "Resized mosaic998_.jpg to (224, 224)\n",
      "Resized mosaic999_.jpg to (224, 224)\n",
      "Resized mosaic99_.jpg to (224, 224)\n",
      "Resized mosaic9_.jpg to (224, 224)\n"
     ]
    }
   ],
   "source": [
    "from PIL import Image\n",
    "import os\n",
    "\n",
    "# Set the input and output directory paths\n",
    "input_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Raw_Data/CCMT_Dataset/Cassava/mosaic'\n",
    "output_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Raw_Data/CCMT_Dataset/Cassava/mosaic'\n",
    "\n",
    "# Ensure the output directory exists, or create it if necessary\n",
    "if not os.path.exists(output_dir):\n",
    "    os.makedirs(output_dir)\n",
    "\n",
    "# Define the target size\n",
    "target_size = (224, 224)\n",
    "\n",
    "# Loop through each image file in the input directory\n",
    "for filename in os.listdir(input_dir):\n",
    "    if filename.endswith(('.jpg', '.jpeg', '.png', '.gif', '.bmp')):  # Add more file extensions if needed\n",
    "        try:\n",
    "            # Open the image file\n",
    "            with Image.open(os.path.join(input_dir, filename)) as img:\n",
    "                # Resize the image to the target size\n",
    "                img_resized = img.resize(target_size)\n",
    "\n",
    "                # Save the resized image to the output directory\n",
    "                output_path = os.path.join(output_dir, filename)\n",
    "                img_resized.save(output_path)\n",
    "\n",
    "                print(f\"Resized {filename} to {target_size}\")\n",
    "        except Exception as e:\n",
    "            print(f\"Error processing {filename}: {str(e)}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32ed25c2",
   "metadata": {},
   "source": [
    "# Maize Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "44f656e1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1667 images belonging to 5 classes.\n",
      "Found 416 images belonging to 5 classes.\n",
      "Epoch 1/10\n",
      "52/52 [==============================] - 214s 4s/step - loss: 1.5275 - accuracy: 0.5425 - val_loss: 0.9206 - val_accuracy: 0.6202\n",
      "Epoch 2/10\n",
      "52/52 [==============================] - 214s 4s/step - loss: 0.7526 - accuracy: 0.7205 - val_loss: 0.8162 - val_accuracy: 0.7091\n",
      "Epoch 3/10\n",
      "52/52 [==============================] - 214s 4s/step - loss: 0.6557 - accuracy: 0.7572 - val_loss: 0.7643 - val_accuracy: 0.6971\n",
      "Epoch 4/10\n",
      "52/52 [==============================] - 215s 4s/step - loss: 0.6341 - accuracy: 0.7596 - val_loss: 0.7686 - val_accuracy: 0.6875\n",
      "Epoch 5/10\n",
      "52/52 [==============================] - 215s 4s/step - loss: 0.6063 - accuracy: 0.7725 - val_loss: 0.7557 - val_accuracy: 0.6995\n",
      "Epoch 6/10\n",
      "52/52 [==============================] - 218s 4s/step - loss: 0.5810 - accuracy: 0.8006 - val_loss: 0.7336 - val_accuracy: 0.7380\n",
      "Epoch 7/10\n",
      "52/52 [==============================] - 220s 4s/step - loss: 0.5637 - accuracy: 0.7847 - val_loss: 0.7264 - val_accuracy: 0.7428\n",
      "Epoch 8/10\n",
      "52/52 [==============================] - 231s 4s/step - loss: 0.5884 - accuracy: 0.7749 - val_loss: 0.7194 - val_accuracy: 0.7139\n",
      "Epoch 9/10\n",
      "52/52 [==============================] - 224s 4s/step - loss: 0.5201 - accuracy: 0.8049 - val_loss: 0.7420 - val_accuracy: 0.7115\n",
      "Epoch 10/10\n",
      "52/52 [==============================] - 227s 4s/step - loss: 0.5704 - accuracy: 0.7835 - val_loss: 0.7222 - val_accuracy: 0.7380\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ABHIJIT DESHPANDE\\anaconda3\\lib\\site-packages\\keras\\src\\engine\\training.py:3000: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
      "  saving_api.save_model(\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.applications import VGG16\n",
    "from tensorflow.keras import models, layers, optimizers\n",
    "\n",
    "# Set the image dimensions and other parameters\n",
    "img_width, img_height = 224, 224\n",
    "batch_size = 32\n",
    "epochs = 10\n",
    "num_classes = 5\n",
    "\n",
    "# Set the data directory\n",
    "data_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Maize'  # Replace with the actual path\n",
    "\n",
    "# Data augmentation and other preprocessing for the training dataset\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1.0 / 255,\n",
    "    shear_range=0.2,\n",
    "    zoom_range=0.2,\n",
    "    horizontal_flip=True,\n",
    "    rotation_range=20,\n",
    "    width_shift_range=0.2,\n",
    "    height_shift_range=0.2,\n",
    "    validation_split=0.2\n",
    ")\n",
    "\n",
    "# Load the training and validation datasets with data augmentation\n",
    "train_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='training'\n",
    ")\n",
    "\n",
    "validation_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation'\n",
    ")\n",
    "\n",
    "# Load the pre-trained VGG16 model without the top (fully connected) layers\n",
    "base_model = VGG16(weights='imagenet', include_top=False, input_shape=(img_width, img_height, 3))\n",
    "\n",
    "# Freeze the layers of the pre-trained VGG16 base to avoid overfitting\n",
    "for layer in base_model.layers:\n",
    "    layer.trainable = False\n",
    "\n",
    "# Add new layers for our specific classification task\n",
    "model = models.Sequential()\n",
    "model.add(base_model)\n",
    "model.add(layers.Flatten())\n",
    "model.add(layers.Dense(256, activation='relu'))\n",
    "model.add(layers.Dropout(0.5))\n",
    "model.add(layers.Dense(num_classes, activation='softmax'))\n",
    "\n",
    "# Compile the model\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "# Train the model with early stopping\n",
    "early_stopping = tf.keras.callbacks.EarlyStopping(patience=3, restore_best_weights=True)\n",
    "history = model.fit(\n",
    "    train_generator,\n",
    "    steps_per_epoch=train_generator.samples // batch_size,\n",
    "    validation_data=validation_generator,\n",
    "    validation_steps=validation_generator.samples // batch_size,\n",
    "    epochs=epochs,\n",
    "    callbacks=[early_stopping]\n",
    ")\n",
    "\n",
    "# Save the model\n",
    "model.save('maize_detection_model.h5')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6feb524c",
   "metadata": {},
   "source": [
    "# Maize Testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "62737ba7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 5943 images belonging to 5 classes.\n",
      "185/185 [==============================] - 659s 4s/step - loss: 0.4988 - accuracy: 0.8304\n",
      "Testing Accuracy: 0.8304054141044617\n",
      "186/186 [==============================] - 719s 4s/step\n",
      "Confusion Matrix:\n",
      "[[ 81  93  46 160 186]\n",
      " [151 198  99 321 315]\n",
      " [ 59  81  41 118 118]\n",
      " [245 356 154 548 579]\n",
      " [275 346 161 603 609]]\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from tensorflow.keras.models import load_model\n",
    "from PIL import Image, ImageFile\n",
    "ImageFile.LOAD_TRUNCATED_IMAGES = True\n",
    "\n",
    "# Load the trained model\n",
    "model = load_model('maize_detection_model.h5')\n",
    "\n",
    "# Load the test dataset with only the \"healthy\" category\n",
    "test_datagen = ImageDataGenerator(rescale=1.0 / 255)\n",
    "test_generator = test_datagen.flow_from_directory(\n",
    "    'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Raw_Data/Health',\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    shuffle=True\n",
    ")\n",
    "\n",
    "# Evaluate the model on the \"healthy\" test dataset\n",
    "loss, accuracy = model.evaluate(\n",
    "    test_generator,\n",
    "    steps=test_generator.samples // batch_size\n",
    ")\n",
    "\n",
    "print(f\"Maize Testing Accuracy: {accuracy}\")\n",
    "\n",
    "# Get the true labels\n",
    "true_labels = test_generator.classes\n",
    "\n",
    "# Get the predicted labels\n",
    "predictions = model.predict(test_generator)\n",
    "predicted_labels = np.argmax(predictions, axis=1)\n",
    "\n",
    "# Compute the confusion matrix\n",
    "cm = confusion_matrix(true_labels, predicted_labels)\n",
    "\n",
    "print(\"Confusion Matrix:\")\n",
    "print(cm)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "91838547",
   "metadata": {},
   "source": [
    "# Tomato training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "9aeb27a4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 13034 images belonging to 5 classes.\n",
      "Found 8689 images belonging to 5 classes.\n",
      "WARNING:tensorflow:`input_shape` is undefined or non-square, or `rows` is not in [96, 128, 160, 192, 224]. Weights for input shape (224, 224) will be loaded as the default.\n",
      "Epoch 1/50\n",
      "407/407 [==============================] - 492s 1s/step - loss: 1.2181 - accuracy: 0.4905 - val_loss: 1.1513 - val_accuracy: 0.5035\n",
      "Epoch 2/50\n",
      "407/407 [==============================] - 307s 754ms/step - loss: 1.1023 - accuracy: 0.5329 - val_loss: 1.1056 - val_accuracy: 0.5167\n",
      "Epoch 3/50\n",
      "407/407 [==============================] - 309s 759ms/step - loss: 1.0569 - accuracy: 0.5507 - val_loss: 1.1122 - val_accuracy: 0.5271\n",
      "Epoch 4/50\n",
      "407/407 [==============================] - 309s 761ms/step - loss: 1.0489 - accuracy: 0.5555 - val_loss: 1.0833 - val_accuracy: 0.5411\n",
      "Epoch 5/50\n",
      "407/407 [==============================] - 308s 757ms/step - loss: 1.0258 - accuracy: 0.5688 - val_loss: 1.0845 - val_accuracy: 0.5366\n",
      "Epoch 6/50\n",
      "407/407 [==============================] - 332s 817ms/step - loss: 1.0106 - accuracy: 0.5788 - val_loss: 1.0919 - val_accuracy: 0.5277\n",
      "Epoch 7/50\n",
      "407/407 [==============================] - 382s 939ms/step - loss: 1.0170 - accuracy: 0.5694 - val_loss: 1.0789 - val_accuracy: 0.5567\n",
      "Epoch 8/50\n",
      "407/407 [==============================] - 601s 1s/step - loss: 1.0030 - accuracy: 0.5801 - val_loss: 1.0824 - val_accuracy: 0.5521\n",
      "Epoch 9/50\n",
      "407/407 [==============================] - 726s 2s/step - loss: 0.9884 - accuracy: 0.5866 - val_loss: 1.0635 - val_accuracy: 0.5658\n",
      "Epoch 10/50\n",
      "407/407 [==============================] - 474s 1s/step - loss: 0.9765 - accuracy: 0.5864 - val_loss: 1.0682 - val_accuracy: 0.5545\n",
      "Epoch 11/50\n",
      "407/407 [==============================] - 463s 1s/step - loss: 0.9726 - accuracy: 0.5902 - val_loss: 1.0739 - val_accuracy: 0.5583\n",
      "Epoch 12/50\n",
      "407/407 [==============================] - 445s 1s/step - loss: 0.9742 - accuracy: 0.5881 - val_loss: 1.0613 - val_accuracy: 0.5506\n",
      "Epoch 13/50\n",
      "407/407 [==============================] - 483s 1s/step - loss: 0.9597 - accuracy: 0.5962 - val_loss: 1.0700 - val_accuracy: 0.5545\n",
      "Epoch 14/50\n",
      "407/407 [==============================] - 546s 1s/step - loss: 0.9636 - accuracy: 0.5934 - val_loss: 1.0780 - val_accuracy: 0.5506\n",
      "Epoch 15/50\n",
      "407/407 [==============================] - 562s 1s/step - loss: 0.9505 - accuracy: 0.5974 - val_loss: 1.0650 - val_accuracy: 0.5512\n",
      "Epoch 16/50\n",
      "407/407 [==============================] - 675s 2s/step - loss: 0.9459 - accuracy: 0.6014 - val_loss: 1.0582 - val_accuracy: 0.5642\n",
      "Epoch 17/50\n",
      "407/407 [==============================] - 859s 2s/step - loss: 0.9386 - accuracy: 0.6051 - val_loss: 1.0571 - val_accuracy: 0.5652\n",
      "Epoch 18/50\n",
      "407/407 [==============================] - 720s 2s/step - loss: 0.9393 - accuracy: 0.5993 - val_loss: 1.0728 - val_accuracy: 0.5646\n",
      "Epoch 19/50\n",
      "407/407 [==============================] - 725s 2s/step - loss: 0.9376 - accuracy: 0.6053 - val_loss: 1.0515 - val_accuracy: 0.5766\n",
      "Epoch 20/50\n",
      "407/407 [==============================] - 487s 1s/step - loss: 0.9254 - accuracy: 0.6107 - val_loss: 1.0664 - val_accuracy: 0.5745\n",
      "Epoch 21/50\n",
      "407/407 [==============================] - 307s 756ms/step - loss: 0.9375 - accuracy: 0.6078 - val_loss: 1.0469 - val_accuracy: 0.5713\n",
      "Epoch 22/50\n",
      "407/407 [==============================] - 307s 755ms/step - loss: 0.9119 - accuracy: 0.6175 - val_loss: 1.0590 - val_accuracy: 0.5625\n",
      "Epoch 23/50\n",
      "407/407 [==============================] - 307s 755ms/step - loss: 0.9250 - accuracy: 0.6087 - val_loss: 1.0645 - val_accuracy: 0.5609\n",
      "Epoch 24/50\n",
      "407/407 [==============================] - 307s 755ms/step - loss: 0.9189 - accuracy: 0.6141 - val_loss: 1.0670 - val_accuracy: 0.5670\n",
      "Epoch 25/50\n",
      "407/407 [==============================] - 308s 756ms/step - loss: 0.9183 - accuracy: 0.6118 - val_loss: 1.0510 - val_accuracy: 0.5717\n",
      "Epoch 26/50\n",
      "407/407 [==============================] - 308s 756ms/step - loss: 0.9116 - accuracy: 0.6168 - val_loss: 1.0733 - val_accuracy: 0.5560\n",
      "Epoch 27/50\n",
      "407/407 [==============================] - 312s 767ms/step - loss: 0.8996 - accuracy: 0.6227 - val_loss: 1.0544 - val_accuracy: 0.5763\n",
      "Epoch 28/50\n",
      "407/407 [==============================] - 324s 797ms/step - loss: 0.9106 - accuracy: 0.6174 - val_loss: 1.0808 - val_accuracy: 0.5495\n",
      "Epoch 29/50\n",
      "407/407 [==============================] - 325s 798ms/step - loss: 0.9078 - accuracy: 0.6162 - val_loss: 1.0804 - val_accuracy: 0.5540\n",
      "Found 8689 images belonging to 5 classes.\n",
      "271/271 [==============================] - 130s 479ms/step - loss: 1.0427 - accuracy: 0.5694\n",
      "Testing Accuracy: 0.5694188475608826\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.applications import MobileNetV2\n",
    "from tensorflow.keras import models, layers, optimizers\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# Set the image dimensions and other parameters\n",
    "img_width, img_height = 200, 200\n",
    "batch_size = 32\n",
    "epochs = 50\n",
    "num_classes = 5\n",
    "\n",
    "# Set the data directory\n",
    "data_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/CCMT_Dataset-Augmented/Tomato/train_set'\n",
    "\n",
    "# Data augmentation and preprocessing for the training dataset\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1.0 / 255,\n",
    "    shear_range=0.2,\n",
    "    zoom_range=0.2,\n",
    "    horizontal_flip=True,\n",
    "    rotation_range=20,\n",
    "    width_shift_range=0.3,\n",
    "    height_shift_range=0.2,\n",
    "    validation_split=0.4\n",
    ")\n",
    "\n",
    "# Load the data and split into training and validation sets\n",
    "data_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='training'\n",
    ")\n",
    "\n",
    "validation_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation'\n",
    ")\n",
    "\n",
    "# Load the MobileNetV2 model without the top (fully connected) layers\n",
    "base_model = tf.keras.applications.MobileNetV2(weights='imagenet', include_top=False, input_shape=(img_width, img_height, 3))\n",
    "\n",
    "# Freeze the layers of the base MobileNetV2 model\n",
    "for layer in base_model.layers:\n",
    "    layer.trainable = False\n",
    "\n",
    "# Add new layers for classification\n",
    "model = models.Sequential()\n",
    "model.add(base_model)\n",
    "model.add(layers.GlobalAveragePooling2D())\n",
    "model.add(layers.Dense(256, activation='relu'))\n",
    "model.add(layers.Dropout(0.5))\n",
    "model.add(layers.Dense(num_classes, activation='softmax'))\n",
    "\n",
    "# Compile the model\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "# Train the model with early stopping\n",
    "early_stopping = tf.keras.callbacks.EarlyStopping(patience=8, restore_best_weights=True)\n",
    "history = model.fit(\n",
    "    data_generator,\n",
    "    steps_per_epoch=data_generator.samples // batch_size,\n",
    "    validation_data=validation_generator,\n",
    "    validation_steps=validation_generator.samples // batch_size,\n",
    "    epochs=epochs,\n",
    "    callbacks=[early_stopping]\n",
    ")\n",
    "\n",
    "# Save the model\n",
    "model.save('Tomato_MobileNet_model.h5')\n",
    "\n",
    "# Evaluate the model on the testing set\n",
    "test_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation',\n",
    "    shuffle=False\n",
    ")\n",
    "\n",
    "loss, accuracy = model.evaluate(\n",
    "    test_generator,\n",
    "    steps=test_generator.samples // batch_size\n",
    ")\n",
    "\n",
    "print(f\"Testing Accuracy: {accuracy}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9dde109d",
   "metadata": {},
   "source": [
    "# Tomato Testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "493d55f8",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 7508 images belonging to 5 classes.\n",
      "234/234 [==============================] - 122s 519ms/step - loss: 0.2476 - accuracy: 0.9061\n",
      "Maize Testing Accuracy: 0.9061164259910583\n",
      "235/235 [==============================] - 123s 519ms/step\n",
      "Confusion Matrix:\n",
      "[[1057  430  304  420  403]\n",
      " [ 586  247  201  226  221]\n",
      " [ 410  177  115  152  161]\n",
      " [ 451  219  150  193  180]\n",
      " [ 447  237  150  188  183]]\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from tensorflow.keras.models import load_model\n",
    "from PIL import Image, ImageFile\n",
    "ImageFile.LOAD_TRUNCATED_IMAGES = True\n",
    "\n",
    "# Load the trained model\n",
    "model = load_model('MobileNet_model.h5')\n",
    "\n",
    "# Load the test dataset with only the \"healthy\" category\n",
    "test_datagen = ImageDataGenerator(rescale=1.0 / 255)\n",
    "test_generator = test_datagen.flow_from_directory(\n",
    "    'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Raw_Data/CCMT_Dataset/Cassava',\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    shuffle=True\n",
    ")\n",
    "\n",
    "# Evaluate the model on the \"healthy\" test dataset\n",
    "loss, accuracy = model.evaluate(\n",
    "    test_generator,\n",
    "    steps=test_generator.samples // batch_size\n",
    ")\n",
    "\n",
    "print(f\"Maize Testing Accuracy: {accuracy}\")\n",
    "\n",
    "# Get the true labels\n",
    "true_labels = test_generator.classes\n",
    "\n",
    "# Get the predicted labels\n",
    "predictions = model.predict(test_generator)\n",
    "predicted_labels = np.argmax(predictions, axis=1)\n",
    "\n",
    "# Compute the confusion matrix\n",
    "cm = confusion_matrix(true_labels, predicted_labels)\n",
    "\n",
    "print(\"Confusion Matrix:\")\n",
    "print(cm)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6161910e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 2630 images belonging to 5 classes.\n",
      "Found 653 images belonging to 5 classes.\n",
      "Epoch 1/10\n",
      "109/109 [==============================] - 1231s 11s/step - loss: 2.8368 - accuracy: 0.2437 - val_loss: 1.6014 - val_accuracy: 0.2639\n",
      "Epoch 2/10\n",
      "109/109 [==============================] - 1222s 11s/step - loss: 1.5992 - accuracy: 0.2824 - val_loss: 1.5898 - val_accuracy: 0.2840\n",
      "Epoch 3/10\n",
      "109/109 [==============================] - 53652s 497s/step - loss: 1.5920 - accuracy: 0.2728 - val_loss: 1.5816 - val_accuracy: 0.2855\n",
      "Epoch 4/10\n",
      "109/109 [==============================] - 1245s 11s/step - loss: 1.5830 - accuracy: 0.2832 - val_loss: 1.5766 - val_accuracy: 0.2840\n",
      "Epoch 5/10\n",
      " 24/109 [=====>........................] - ETA: 12:21 - loss: 1.5791 - accuracy: 0.2951"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.applications import VGG16\n",
    "from tensorflow.keras import models, layers, optimizers\n",
    "\n",
    "# Set the image dimensions and other parameters\n",
    "img_width, img_height = 400, 400\n",
    "batch_size = 24\n",
    "epochs = 10\n",
    "num_classes = 5\n",
    "\n",
    "# Set the data directory\n",
    "data_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Check/tomato'\n",
    "\n",
    "# Data augmentation and other preprocessing for the training dataset\n",
    "train_datagen = ImageDataGenerator(\n",
    "    rescale=1.0 / 255,\n",
    "    shear_range=0.2,\n",
    "    zoom_range=0.2,\n",
    "    horizontal_flip=True,\n",
    "    rotation_range=20,       # Additional data augmentation: rotation\n",
    "    width_shift_range=0.2,   # Additional data augmentation: horizontal shift\n",
    "    height_shift_range=0.2,  # Additional data augmentation: vertical shift\n",
    "    validation_split=0.2     # Split data into training and validation sets\n",
    ")\n",
    "\n",
    "# Load the training and validation datasets with data augmentation\n",
    "train_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='training'\n",
    ")\n",
    "\n",
    "validation_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation'\n",
    ")\n",
    "\n",
    "# Load the pre-trained VGG16 model without the top (fully connected) layers\n",
    "base_model = VGG16(weights='imagenet', include_top=False, input_shape=(img_width, img_height, 3))\n",
    "\n",
    "# Freeze the layers of the pre-trained VGG16 base to avoid overfitting\n",
    "for layer in base_model.layers:\n",
    "    layer.trainable = False\n",
    "\n",
    "# Add new layers for our specific classification task\n",
    "model = models.Sequential()\n",
    "model.add(base_model)\n",
    "model.add(layers.Flatten())\n",
    "model.add(layers.Dense(256, activation='relu'))\n",
    "model.add(layers.Dropout(0.5))\n",
    "model.add(layers.Dense(num_classes, activation='softmax'))\n",
    "\n",
    "# Compile the model\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "# Train the model with early stopping\n",
    "early_stopping = tf.keras.callbacks.EarlyStopping(patience=3, restore_best_weights=True)\n",
    "history = model.fit(\n",
    "    train_generator,\n",
    "    steps_per_epoch=train_generator.samples // batch_size,\n",
    "    validation_data=validation_generator,\n",
    "    validation_steps=validation_generator.samples // batch_size,\n",
    "    epochs=epochs,\n",
    "    callbacks=[early_stopping]\n",
    ")\n",
    "\n",
    "# Save the model\n",
    "model.save('tomato_model.h5')\n",
    "\n",
    "# Evaluate the model on the testing set\n",
    "test_generator = train_datagen.flow_from_directory(\n",
    "    data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    subset='validation',\n",
    "    shuffle=False\n",
    ")\n",
    "\n",
    "loss, accuracy = model.evaluate(\n",
    "    test_generator,\n",
    "    steps=test_generator.samples // batch_size\n",
    ")\n",
    "\n",
    "print(f\"Tomato Testing Accuracy: {accuracy}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f1f70494",
   "metadata": {},
   "source": [
    "# Tomato testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "d5932e0f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 5445 images belonging to 5 classes.\n",
      "170/170 [==============================] - 608s 4s/step - loss: 5.2640 - accuracy: 0.1147\n",
      "Tomato Testing Accuracy: 0.1147058829665184\n",
      "171/171 [==============================] - 633s 4s/step\n",
      "Confusion Matrix:\n",
      "[[ 177   16    7    9  291]\n",
      " [ 463   54   26   16  750]\n",
      " [ 186   15   13    6  312]\n",
      " [ 893   66   37   26 1318]\n",
      " [ 279   28   15   12  430]]\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from tensorflow.keras.models import load_model\n",
    "from PIL import Image, ImageFile\n",
    "ImageFile.LOAD_TRUNCATED_IMAGES = True\n",
    "\n",
    "# Load the trained model\n",
    "model = load_model('maize_detection_model.h5')\n",
    "\n",
    "# Load the test dataset with only the \"healthy\" category\n",
    "test_datagen = ImageDataGenerator(rescale=1.0 / 255)\n",
    "test_generator = test_datagen.flow_from_directory(\n",
    "    'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/CCMT_Dataset-Augmented/Tomato/test_set',\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    shuffle=True\n",
    ")\n",
    "\n",
    "# Evaluate the model on the \"healthy\" test dataset\n",
    "loss, accuracy = model.evaluate(\n",
    "    test_generator,\n",
    "    steps=test_generator.samples // batch_size\n",
    ")\n",
    "\n",
    "print(f\"Tomato Testing Accuracy: {accuracy}\")\n",
    "\n",
    "# Get the true labels\n",
    "true_labels = test_generator.classes\n",
    "\n",
    "# Get the predicted labels\n",
    "predictions = model.predict(test_generator)\n",
    "predicted_labels = np.argmax(predictions, axis=1)\n",
    "\n",
    "# Compute the confusion matrix\n",
    "cm = confusion_matrix(true_labels, predicted_labels)\n",
    "\n",
    "print(\"Confusion Matrix:\")\n",
    "print(cm)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36c67c79",
   "metadata": {},
   "source": [
    "# Plotting the graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "3d0d2aa1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.applications import EfficientNetB0\n",
    "from tensorflow.keras import layers, models\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Plot training and validation loss\n",
    "plt.plot(history.history['loss'], label='Training Loss', color='red')\n",
    "plt.plot(history.history['val_loss'], label='Validation Loss', color='blue')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# Plot training and validation accuracy\n",
    "plt.plot(history.history['accuracy'], label='Training Accuracy', color='red')\n",
    "plt.plot(history.history['val_accuracy'], label='Validation Accuracy', color='blue')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "834e6694",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from PIL import Image\n",
    "\n",
    "def is_valid_image(image_path):\n",
    "    try:\n",
    "        with Image.open(image_path) as img:\n",
    "            img.verify()\n",
    "            return True\n",
    "    except Exception as e:\n",
    "        print(f\"Error: {e} for image: {image_path}\")\n",
    "        return False\n",
    "\n",
    "def remove_broken_images(data_dir):\n",
    "    for root, _, files in os.walk(data_dir):\n",
    "        for file in files:\n",
    "            image_path = os.path.join(root, file)\n",
    "            if not is_valid_image(image_path):\n",
    "                os.remove(image_path)\n",
    "                print(f\"Removed {image_path} (broken image)\")\n",
    "\n",
    "# Set the data directory\n",
    "data_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Raw_Data/Tomato'  # Replace with the path to your data directory\n",
    "remove_broken_images(data_dir)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "63ec056e",
   "metadata": {},
   "source": [
    "# Confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "ac257ee5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 5943 images belonging to 5 classes.\n",
      "186/186 [==============================] - 1299s 7s/step\n",
      "[4 4 3 ... 4 3 2]\n",
      "[0 0 0 ... 4 4 4]\n",
      "Confusion Matrix:\n",
      "[[ 77 103  44 174 168]\n",
      " [139 185  79 319 362]\n",
      " [ 47  77  35 115 143]\n",
      " [281 341 183 547 530]\n",
      " [267 368 160 595 604]]\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from tensorflow.keras.models import load_model\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "# Set the image dimensions and other parameters\n",
    "img_width, img_height = 224, 224\n",
    "batch_size = 32\n",
    "\n",
    "# Set the data directory for the 2 classes you want to test on\n",
    "test_data_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Raw_Data/Health'  # Replace with the path to the new test directory\n",
    "\n",
    "# Load the trained model\n",
    "model = load_model('maize_detection_model.h5')  # Replace with the path to your trained model\n",
    "\n",
    "# Create the test data generator for the 2 classes\n",
    "test_datagen = ImageDataGenerator(rescale=1.0 / 255)\n",
    "test_generator = test_datagen.flow_from_directory(\n",
    "    test_data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    shuffle=True\n",
    ")\n",
    "\n",
    "# Evaluate the model on the new test dataset\n",
    "predictions = model.predict(test_generator)\n",
    "predicted_classes = np.argmax(predictions, axis=1)\n",
    "true_classes = test_generator.classes\n",
    "\n",
    "# Compute confusion matrix\n",
    "confusion = confusion_matrix(true_classes, predicted_classes)\n",
    "\n",
    "print(predicted_classes)\n",
    "print(true_classes)\n",
    "# Calculate accuracy manually\n",
    "#correct_predictions = np.sum(predicted_classes == true_classes)\n",
    "#total_samples = len(true_classes)\n",
    "#accuracy = correct_predictions / total_samples\n",
    "\n",
    "print(\"Confusion Matrix:\")\n",
    "print(confusion)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "db63fdc1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.09      0.14      0.11       566\n",
      "           1       0.17      0.17      0.17      1084\n",
      "           2       0.07      0.08      0.08       417\n",
      "           3       0.31      0.29      0.30      1882\n",
      "           4       0.33      0.30      0.32      1994\n",
      "\n",
      "    accuracy                           0.24      5943\n",
      "   macro avg       0.20      0.20      0.20      5943\n",
      "weighted avg       0.26      0.24      0.25      5943\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "\n",
    "# Calculate precision, recall, and F1-score for each class\n",
    "report = classification_report(true_classes, predicted_classes)\n",
    "print(report)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3938ad87",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score\n",
    "\n",
    "# Assuming true_classes and predicted_classes are the true and predicted class labels\n",
    "# true_classes = test_generator.classes\n",
    "# predicted_classes = model.predict_classes(test_generator)\n",
    "\n",
    "# Choose the class index for which you want to assess accuracy\n",
    "target_class = 4 # Replace with the index of the class you want to evaluate\n",
    "\n",
    "# Create a binary vector indicating if the true class matches the target class\n",
    "true_target_class = np.where(true_classes == target_class, 1, 0)\n",
    "predicted_target_class = np.where(predicted_classes == target_class, 1, 0)\n",
    "\n",
    "# Compute confusion matrix for the target class\n",
    "confusion_target = confusion_matrix(true_target_class, predicted_target_class)\n",
    "\n",
    "# Compute accuracy for the target class\n",
    "accuracy_target = accuracy_score(true_target_class, predicted_target_class)\n",
    "\n",
    "print(\"Confusion Matrix for Target Class:\")\n",
    "print(confusion_target)\n",
    "\n",
    "print(f\"Accuracy for Target Class: {accuracy_target}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "371c8436",
   "metadata": {},
   "source": [
    "# VGG19 \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b50f9ebd",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dir = 'F:/College/Years/B.Tech/Major_project/Datasets/Dataset_for_Crop_Pest_and_Disease_Detection/Raw_Data/Tomato'  # Replace with the path to your dataset directory\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fd48a8a6",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}