Histogram matching is a process where an image is altered in such a way that its histogram matches that of another reference (template) image's histogram. The algorithm is as follows:
- The cumulative histogram is computed for each image, as shown in the following screenshot.
- For any given pixel value, xi, in the input image (to be adjusted), we need to find the corresponding pixel value, xj, in the output image by matching the input image's histogram with the template image's histogram.
- The xi pixel value has a cumulative histogram value given by G(xi). Find a pixel value, xj,so that the cumulative distribution value in the reference image, namely H(xj), is equal to G(xi).
- The input data value xi is replaced by xj:
