How do I augment images using keras ImageDataGenerator?

Question

I have training set of 364 images stored in numpy array also its labels stored in a different numpy array(there are 8 labels to classify).The dataset being small I want to use augmentation but can only find resources which augment images if they are stored in specific folders according to the labels. So how can I augment images in real time using ImageDataGenerator. Thanks!

Answer 1

Solution

You should use flow instead of flow_from_directory. flow can take your numpy arrays. I am pasting code from coursera course.

training_images = training_images.reshape((27455,28,28,1))
testing_images = testing_images.reshape((7172,28,28,1))

# Create an ImageDataGenerator and do Image Augmentation
training_datagen = ImageDataGenerator(
      rescale = 1./255,
      rotation_range=40,
      width_shift_range=0.2,
      height_shift_range=0.2,
      shear_range=0.2,
      zoom_range=0.2,
      horizontal_flip=True,
      fill_mode='nearest')

validation_datagen = ImageDataGenerator(rescale = 1./255)

training_datagen.fit(training_images)
validation_datagen.fit(testing_images)

testing_labels = tf.one_hot(testing_labels, 25)
training_labels = tf.one_hot(training_labels, 25)

history = model.fit_generator(training_datagen.flow(
    training_images,
    y = training_labels,
    batch_size = 32
), epochs=15, validation_data = validation_datagen.flow(
    testing_images,
    y = testing_labels,
    batch_size = 32
), verbose = 1)

Edit 1

Name of Coursera Course: Convolutional Neural Networks in TensorFlow by deeplearning.ai

How to know images increased or not

fix batch size and see how many steps training takes to complete 1 epoch. From that you can infer, images are augmenting.

I believe this solution solves your problem. If yes accept it else comment below what's the problem?

Edit 2

# here's a more "manual" example
for e in range(epochs):
    print('Epoch', e)
    batches = 0
    for x_batch, y_batch in training_datagen.flow(x_train, y_train, batch_size=32):
        ##### Check each batch manually####
        batches += 1

Augmenting particular images

check below article which will show you hoe to augment images one by one.

https://towardsdatascience.com/data-augmentation-techniques-in-python-f216ef5eed69