preprocess_img.py

from cv2 import data
import numpy as np
import cv2
import os
import random
from data_processing import resize

chars = ['A','B','C','D']
train_database = []
test_database = []
for c in chars:
    for path in os.listdir(f'data/{c}'):
        if path == '.DS_Store':
            continue
        for img_path in os.listdir(f'data/{c}/{path}'):
            if img_path == '.DS_Store':
                continue
            image = cv2.imread(f'data/{c}/{path}/{img_path}')
            # scales/pads the images so that they're all shape (100, 100, 3):
            image = resize(image)
            train_database.append([image,ord(path)-ord('a')])

for path in os.listdir(f'data/E'):
    if path == '.DS_Store':
        continue
    for img_path in os.listdir(f'data/{c}/{path}'):
        if img_path == '.DS_Store':
            continue
        image = cv2.imread(f'data/{c}/{path}/{img_path}')
        # scales/pads the images so that they're all shape (100, 100, 3):
        image = resize(image)
        test_database.append([image,ord(path)-ord('a')])

random.shuffle(train_database)
random.shuffle(test_database)

imgs = []
labs = []
for img,label in train_database:
    imgs.append(img)
    labs.append(label)
t_imgs = []
t_labs = []
for img,label in test_database:
    t_imgs.append(img)
    t_labs.append(label)

imgs = np.array(imgs)
labels = np.array(labs)
t_imgs = np.array(t_imgs)
t_labels = np.array(t_labs)
np.save('data/images', imgs)
np.save('data/labels', labels)
np.save('data/test_images', t_imgs)
np.save('data/test_labels', t_labels)