crop

Usage

Crops a region from an image following the specified coordinates of the upper left corner, width, and height of the region. It does not involve image scaling and only extracts the specified region.

Interface

torchvision.transforms.crop(img, top, left, height, width)

Parameters

Parameter	Description	Value Range	Input/Output
img	Input image.	torch.Tensor	Input
top	Upper coordinate of the region.	[0, image height)	Input
left	Left coordinate of the region.	[0, image width)	Input
height	Height of the region.	(0, image height – top]	Input
width	Width of the region.	(0, image width – left]	Input

Return Value

Cropped image

Example

from torchvision import transforms
from PIL import Image

# Open an image.
img = Image.open('example.jpg')

# Define cropping parameters.
top = 50
left = 50
height = 100
width = 100

# Crop the image.
cropped_img = transforms.functional.crop(img, top, left, height, width)

# Save the cropped image.
cropped_img.save('cropped_example.jpg')

Parent topic: Interface Definition