Ask Your Question

Revision history [back]

Individual Channel Gain (convertTo for channels)

I Would like to know how to adjust the "gain" of the individual R, G or B channels in an image.

That is, I have a bunch of images taken from a bunch of individual "gadgets", each gadget uses the same make and model image sensor, make and model lens and make and model lighting arrangement. The gadgets are completely enclosed. The subject of the images are just various solid color backgrounds (yellowish, orangish, bluish, etc.).

Given the controlled conditions, ideally the color variability from one gadget to another would be zero, but in reality the coloring varies slightly. I have analyzed the images and found that for a given gadget the individual channels will vary consistently up or down regardless of which colored background was used compared to the average. For example for gadget "A" the red channel will be consistently higher by some percentage compared to the average of all the gadgets.

What I want to do is adjust the individual channels of individual gadgets so that the bitmap values are more consistent. Is there an efficient way to do this in OpenCV?