Publication: Local contrast enhancement methods based on modified histogram equalization
Date
2022-03-01
Authors
Majeed, Samer Hameed
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Images with poor contrast might be acquired under some circumstances, such as poor capturing environment and insufficient illumination. These captured images could be a significant challenge to computer vision researchers especially the ones with poor contrast. The conventional contrast enhancement techniques including variants of histogram equalization (HE) techniques have several main drawbacks such as manual parameter adjustment, introduction of artifacts and noises, corruption of image’s details, insufficient illumination enhancement, and formation of over enhanced pixels and regions. To reduce these limitations, this study proposes two new HE-based contrast enhancement techniques, which are Iterated Adaptive Entropy-Clip Limit Histogram Equalization (IAECHE) and Adaptive Entropy Index Histogram Equalization (AEIHE). The proposed IAECHE technique divides the image into multi sub-images based on the dimensions of an input image. Each sub-image will then be divided into contextual regions before individually enhanced using the conventional Contrast Limited Adaptive Histogram Equalization (CLAHE). Different to CLAHE, the value of the clip limit is adaptively and automatically set. On the other hand, the proposed AEIHE technique divides the input image into three sub-images. Then, the sub-images will be enhanced individually by applying an adaptive and automatic window size with adaptive and automatic clip limit depending on the distribution of the pixels over the gray levels of these sub-images. The optimum values of the window size and the clip limit are obtained by combining the Whale Optimization Algorithm (WOA), Grasshopper Optimization Algorithm (GOA), and Particle Swarm Optimization (PSO) with the conventional CLAHE. Both techniques have been compared with 10 state-of-the-art HE-based contrast enhancement techniques. For qualitative analysis, the proposed IAECHE has produced better enhanced images by improving the local contrast and highlighting the local details of the resultant images, while the proposed AEIHE has successfully improved the images’ local contrast, preserved the images’ structure, and highlighted the images’ hidden details. This qualitative analysis has been supported by quantitative analysis. The results strongly indicate that the proposed IAECHE and AEIHE techniques could possibly be used to pre-process images in many applications such as surveillance, medical and security purposes.