Histogram equalization is a technique used to improve the visual appearance of digital images and amplify smaller details. It involves dividing an image into three different color channels and creating a histogram to represent the frequency of occurrence of each intensity level. A cumulative distribution is then created, which can be normalized and scaled to the range of [0,255]. Each pixel in the original image is then replaced with the new corresponding value, and this operation can be done on each color channel. Block histogram equalization is another technique that involves dividing an image into different blocks and applying histogram equalization separately to each region. This technique can be done on non-overlapping regions, and the transfer functions can be smoothly interpolated to avoid blocking artifacts and reduce noise amplification. There are many variations and implementations of adaptive histogram equalization techniques.

source update: Enhancing the Local Details With Adaptive Histogram Equalization – Towards AI


There are no comments yet.

Leave a comment