giladamar / new_histogram_equalization

Implementation of the paper https://arxiv.org/pdf/2101.05922v1.pdf

This repo is an implementation of the paper https://arxiv.org/pdf/2101.05922v1.pdf in which a new algorithm for grayscale histogram equalization to improve image contrast.

I have found that in some instances the performance is, visually, better than that of OpenCV's histogram equalization. However, the result I get for the camera man imagee, includes in _ examples_ is not nearly as good as the paper shows. This may be an issue with my implementation of the paper of course.

The requirement is just numpy in order to use the algorithm contained in _ Improved_Equalization.py_ , but the notebook which illustrates the algorithm details requires openCV as well.

Feel free to make an issue, PR, or send me a mail if you have any questions.