Time to dive right in! Image augmentation using imgaug is as easy as importing the library and doing some basic preparations in your Python code:

import imgaug as ia
from imgaug import augmenters as iaa
import numpy as np

# This seed can be changed - random seed
ia.seed(1)

# Example batch of 100 images
images = np.array(
    [ia.quokka(size=(64, 64)) for _ in range(100)],
    dtype=np.uint8
)

This step essentially allows us to use the imgaug library as iaa and import some demo images to augment to show the power of image augmentation. Next, we are going to create a transformer function to specify all of the different augmentations that we'd like applied to our images:

# Create the transformer function by specifying the different augmentations
seq = iaa.Sequential([
    # Horizontal Flips
    iaa.Fliplr(0.5), 

    # Random Crops
    iaa.Crop(percent=(0, 0.1)), 

    # Gaussian blur for 50% of the images
    iaa.Sometimes(0.5,
        iaa.GaussianBlur(sigma=(0, 0.5))
    ),
    # Strengthen or weaken the contrast in each image.
    iaa.ContrastNormalization((0.75, 1.5)),

    # Add gaussian noise.
    iaa.AdditiveGaussianNoise(loc=0, scale=(0.0, 0.05*255), per_channel=0.5),

    # Make some images brighter and some darker.
    iaa.Multiply((0.8, 1.2), per_channel=0.2),

    # Apply affine transformations to each image.
    iaa.Affine(
        scale={"x": (0.5, 1.5), "y": (0.5, 1.5)},
        translate_percent={"x": (-0.5, 0.5), "y": (-0.5, 0.5)},
        rotate=(-10, 10),
        shear=(-10, 10)
    )
], 
# apply augmenters in random order
random_order=True)

You'll notice there are quite a few different augmentations included in this demo piece. One of the great things about this library is that it offers a wide assortment of ready to use augmentations that're compatible with deep learning and flexible to be used in other applications. Finally, we need to display the images to a window to understand what types of augmentations we've applied to this images:

# This should display a random set of augmentations in a window
images_aug = seq.augment_images(images)
seq.show_grid(images[0], cols=8, rows=8

I encourage you to experiment with the different types of augmentations when you get to the exercise section in this chapter.