I have two 8bit unsigned RGBA images, and I am trying to blend them together according to their alpha transparencies , in Python, using cv2.cuda.alphacomp
I notice that the pixels which get blended have the wrong colour
For example, using ALPHA_OVER results in the darkening of the pixel, while using ALPHA_PLUS results in a brightening of the pixel
is there a way to avoid these issues?
Can you provide me with an MRE and the documentation explaining the functionality of ALPHA_OVER and ALPHA_PLUS and I’ll take a look.