Topic: Histogram Matching
I use IA for processing satellite imagery of the Earth and for this purpose it is ideal having many features found only in otherwise very expensive software.
However, it would be very useful if IA were able to perform a few more very useful image processing transforms, one of these is the CDF reference stretch.
In simple terms it is the matching of the histogram of one image to that of another, so that the two images have a similar colour and contrast, CDF being the cumulative distribution of function of the reference image that a second or more images are to have their histograms matched to.
In Photoshop I think this process is termed "colour matching".
Technically, histogram matching is achieved by forward mapping the image DNs through the image CDF and backmapping through the reference image CDF. The first transformation is the same as histogram equalisation, the second transformation is the inverse of HE using a CDF of the reference image.
In remote sensing this transform is used to compensate for the effects of differing solar illumination or vegetation growth changes evident on images collected at different times, that should appear similar. However, I would imagine this transform would have wide application outside the field of remote sensing, for example in scientific and medical imaging, or general digital photography where it is desired to combine two or more images to make a mosaic (and where lightng differences can prevent this being done well).
Many thanks for considering this suggestion.