From each image block of size 32*32, how to estimate the noise variation using principal component analysis?
To distinguish the forged section of the forged image, I converted the image into HSV first, extracted the saturation component, segmented the image into non-overlapping blocks of size 32*32, then from each block I need to estimate the noise variation using principal component analysis. How to do this? Python code for this will be very helpful. Thank you.