Hi,
I want to implement exposure fusion on black/white images (1 channel 8 bits). The OpenCV documentation has no channel specification on the input images.
So far, The code runs fine, so I guess 1 channel is supportive. However, the original paper and code by Mertens use 3 channel input. I wonder how is each measure implemented in OpenCV with 1 channel input?
For contrast measure, is the conversion to 1 channel skipped?
For saturation measure, is its weight just 0, meaning no effect of this measure? Because there is no standard deviation between channels. We have only 1.
For well-exposedness measure, gauss curve applied to just 1 channel?
I cannot find anything related on the web. Any responses would be appreciated!