Image Processing: Whitening implementation not works as expected

Question

I take the mathematical expression from Computer Vision: Models, Learning and Inference book. Where am i wrong here?

import cv2

grayImage = c