Data Preprocessing for CNNs

image_generators_colab.py

from tensorflow.keras.preprocessing.image import ImageDataGenerator

# Create data generators
# File path = path to train/test/val folders respectively
# Use a target size of 224x224 px for each image (or whatever size you choose)
# Batch size = total number of images in the train set, test set, val set respectively
# Ensure class_mode is binary
train_generator = ImageDataGenerator(rescale=1./255).flow_from_directory(
                                                    '/content/chest_xray/train',
                                                     target_size=(224, 224),
                                                     batch_size=5216,
                                                     class_mode='binary',
                                                     seed=123)
test_generator = ImageDataGenerator(rescale=1./255).flow_from_directory(
                                                    '/content/chest_xray/test',
                                                     target_size=(224, 224),
                                                     batch_size=624,
                                                     class_mode='binary',
                                                     seed=123)
val_generator = ImageDataGenerator(rescale=1./255).flow_from_directory(
                                                    '/content/chest_xray/val',
                                                     target_size=(224, 224),
                                                     batch_size=16,
                                                     class_mode='binary',
                                                     seed=123)

# Iterate through each generator to create the data sets with the train/test/val splits
X_train, y_train = next(train_generator)
X_test, y_test = next(test_generator)
X_val, y_val = next(val_generator)

# Check which class is which
train_generator.class_indices