imgaug_sequential_pipeline_flip_blur_crop_augmentation.py

1import numpy as np
2import imgaug as ia
3import imgaug.augmenters as iaa
4
5# Standardize random seed
6ia.seed(1)
7
8# Example images (4 images of size 64x64 with 3 color channels)
9images = np.array(
10    [ia.quokka(size=(64, 64)) for _ in range(32)],
11    dtype=np.uint8
12)
13
14# Define the augmentation pipeline
15seq = iaa.Sequential([
16    iaa.Fliplr(0.5), # horizontally flip 50% of the images
17    iaa.GaussianBlur(sigma=(0, 3.0)), # blur images with a sigma between 0 and 3.0
18    iaa.Crop(px=(0, 16)) # crop images from each side by 0 to 16px (randomly chosen)
19], random_order=True) # apply augmenters in random order
20
21# Augment the images
22images_aug = seq(images=images)
23
24# Note: In a real scenario, you would now train your model using 'images_aug'