Land Use / Land Cover Classification using EuroSAT

📌 Project Overview

This project performs Land Use / Land Cover (LULC) classification using the EuroSAT satellite image dataset.

Three different models are implemented and compared:

1. Support Vector Machine (SVM)
2. Custom Convolutional Neural Network (CNN)
3. Transfer Learning using ResNet50

📊 Model Performance

Model	Accuracy
SVM	~70%
CNN	~80%
ResNet50 (Transfer Learning)	~96%

📂 Dataset

• Dataset used: EuroSAT
• Download from Kaggle:
  https://www.kaggle.com/datasets/apollo2506/eurosat-dataset

After downloading, extract the dataset so that class folders (AnnualCrop, Forest, etc.) are available.

🧠 Technologies Used

• Python
• OpenCV
• NumPy
• Scikit-learn
• TensorFlow / Keras
• Matplotlib / Seaborn