How to crop an image from the center with certain dimensions?

Question

Say we have an image with the following dimensions:

width = 200
height = 100

Say that the dimensions of our crop would be 50x50.

How can I crop an image using this new dimension from the center of the image, using Python?

Answer 1

Assume a crop box with the dimensions: cw, ch = 50, 50.

The center of the image is the point (w//2, h//2), where w is its width and h is its height. The square crop box which is 50 pixels on a side would also be centered there.

This means the upper left corner of the crop box would be at (w//2 - cw//2, h//2 - ch//2), and its lower right corner is at (w//2 + cw//2, h//2 + ch//2).

There's at least two ways to crop the image I can think of. The first is to use the Image.crop() method and pass it the coordinates of the rectangular crop region.

box = w//2 - cw//2, h//2 - ch//2, w//2 + cw//2, h//2 + ch//2
cropped_img = img.crop(box)

Which can be simplified mathematically to cut down on the number of divisions:

box = (w-cw)//2, (h-ch)//2, (w+cw)//2, (h+ch)//2  # left, upper, right, lower
cropped_img = img.crop(box)

Another way to do it is via the ImageOps.crop() function which is passed the size of the border on each of the four sides:

from PIL import ImageOps

wdif, hdif = (w-cw)//2, (h-ch)//2
border = wdif, hdif, wdif, hdif  # left, top, right, bottom
cropped_img = ImageOps.crop(img, border)