datatSetGenerator.py

import numpy as np
import cv2
from PIL import Image as im
import pandas as pd
import mediapipe as mp
import os
mp_drawing = mp.solutions.drawing_utils
mp_drawing_styles = mp.solutions.drawing_styles
mp_hands = mp.solutions.hands

'''
images = pd.read_csv("sign_mnist_train.csv")

images_numpy = pd.DataFrame.to_numpy(images)
for image in images_numpy:
    name = image[0]
    image_arr = (np.array(image[1:785])).reshape(28, 28)
    print(image_arr)
    rgb_img = np.array([[[0, 0, 0] for i in range(0,28)] for j in range(0,28)])
    for i in range(0, 28):
        for j in range(0, 28):
#            print(image_arr[i][j])
            rgb_img[i][j] = [image_arr[i][j], image_arr[i][j], image_arr[i][j]]
#    detector = handDetector()
    img = im.fromarray(rgb_img.astype(np.uint8))
    img.save("test.png")
#    markedImg = detector.findHands(img)
#    markedImg("test2.png")
    break
'''

IMAGE_FILE_CONTAINER = []
for i in range(10):
    IMAGE_FILE_CONTAINER.append([])
    for filename in os.listdir("./" + str(i)):
        IMAGE_FILE_CONTAINER[i].append("./" + str(i) + "/" + filename)

import csv

# open the file in the write mode
f = open('./data.csv', 'w')
f.write("NUMBER, LANDMARKS\n")

#IMAGE_FILE_CONTAINER = [['IMG_1140.jfif', './0/IMG_1118.JPG']]
with mp_hands.Hands(
    static_image_mode=True,
    max_num_hands=2,
    min_detection_confidence=0.5) as hands:
  for i in range(len(IMAGE_FILE_CONTAINER)):
      for file in IMAGE_FILE_CONTAINER[i]:
        print('file:' + file)
        # Read an image, flip it around y-axis for correct handedness output (see
        # above).
        image = cv2.flip(cv2.imread(file), 1)
        # Convert the BGR image to RGB before processing.
        results = hands.process(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))

        # Print handedness and draw hand landmarks on the image.
        #print('Handedness:', results.multi_handedness)
        if not results.multi_hand_landmarks:
          continue
        image_height, image_width, _ = image.shape
        annotated_image = image.copy()
        for hand_landmarks in results.multi_hand_landmarks:
          f.write(str(i)+',\"'+str(hand_landmarks)+"\"\n")
          #land = 'hand_landmarks:' + '   ' + hand_landmarks + '     '
          #print('hand_landmarks:', hand_landmarks)
          #print(
          #    f'Index finger tip coordinates: (',
          #    f'{hand_landmarks.landmark[mp_hands.HandLandmark.INDEX_FINGER_TIP].x * image_width}, '
          #    f'{hand_landmarks.landmark[mp_hands.HandLandmark.INDEX_FINGER_TIP].y * image_height})'
          #)
          mp_drawing.draw_landmarks(
              annotated_image,
              hand_landmarks,
              mp_hands.HAND_CONNECTIONS,
              mp_drawing_styles.get_default_hand_landmarks_style(),
              mp_drawing_styles.get_default_hand_connections_style())
        #im.fromarray(annotated_image.astype(np.uint8)).save('test2.png')

# close the file
f.close()