Dividing image into patches and getting patch index of one pixel

Question

I have a matrix of size 1080 x 667 (e.g. representing an image). Now I would like to divide this matrix into quadratic non-overlaping patches row-wise and when a (x,y) coordinate is provided (i.e. one pixel in the image), I need the patch number. I thought about a method with the declaration:

def get_patch_index(image_dim, patch_dim, x, y)

image_dim is the dimension of the matrix (1080, 667). patch_dim is the patch size (e.g. 3,3) and x and y is the pixel to evaluate.

For example, a call get_patch_index((1080,667), (3,3), 0, 0) should return an index 0 (the first patch) and a call get_patch_index((1080,667), (3,3), 1080, 667) should return an index 39 (the last patch).

How is it possible to efficiently implement this?

Answer 1

You can use pil library of python. Then follow steps

1. Create matrices of patch dimension and loop through pixel
2. Put the result in order in some list or map
3. write method that has input of dimensions and searches through the map to return patch number