Transformations

Image Processing in Python

Rebeca Gonzalez

Data Engineer

Why transform images?

Preparing images for classification Machine Learning models
Optimization and compression of images
Save images with same proportion

A cat picture being processed, downscaled

Rotating

2 dogs playing, walking

Photo of 2 dogs playing, walking rotated 90 degrees clockwise

Rotating

2 dogs playing, walking

2 dogs playing, walking rotated 90 degreed anticlockwise

Rotating clockwise

from skimage.transform import rotate


# Rotate the image 90 degrees clockwise 
image_rotated = rotate(image, -90)


show_image(image, 'Original')
show_image(image_rotated, 'Rotated 90 degrees clockwise')

Original photo of dogs next to rotated one

Rotating anticlockwise

from skimage.transform import rotate


# Rotate an image 90 degrees anticlockwise
image_rotated = rotate(image, 90)


show_image(image, 'Original')
show_image(image_rotated, 'Rotated 90 degrees anticlockwise')

Original photo of dogs next to 90 degrees rotated one

Rescaling

Original image of 2 dogs with original size of 3000 pixels next to downscaled one with 800 pixels

Rescaling

Downgrading

from skimage.transform import rescale


# Rescale the image to be 4 times smaller 
image_rescaled = rescale(image, 1/4, anti_aliasing=True, multichannel=True)


show_image(image, 'Original image')
show_image(image_rescaled, 'Rescaled image')

Rescaling

Original image of 2 dogs with original size of 3000 pixels next to downscaled one with 800 pixels

Aliasing in digital images

Letter A image without aliasing next an image of letter A with aliasing

Aliasing in digital images

Rescaled image of 2 dogs playing with aliasing next to a rescaled one without aliasing

Resizing

Original image with 3000 pixels, next to a resized one with 800 pixels

Resizing

from skimage.transform import resize


# Height and width to resize
height = 400
width = 500


# Resize image
image_resized = resize(image, (height, width), anti_aliasing=True)


# Show the original and resulting images
show_image(image, 'Original image')
show_image(image_resized, 'Resized image')

Resizing

Original image of 2 dogs with original size of 3000 pixels next to resized one, not looking proportional

Resizing proportionally

from skimage.transform import resize


# Set proportional height so its 4 times its size
height = image.shape[0] / 4
width = image.shape[1] / 4


# Resize image
image_resized = resize(image, (height, width), anti_aliasing=True)


show_image(image_resized, 'Resized image')

Resizing proportionally

Original image of 2 dogs with original size of 3000 pixels next to a resized one, this time proportionally with 800 pixels

Let's practice!

Image Processing in Python