Image Contrast Enhancement Algorithm Based on Discriminant Analysis

An image contrast enhancement algorithm based on discriminant analysis is presented in this paper. An automatic adaptive segmentation algorithm, based on discriminant analysis, is utilized to segment the gray-values of an image into several clusters. After the adaptive segmentation algorithm, different object components of the image are segmented into separate clusters, called object planes. Then, the dynamic range of each object plane is adjusted according to its statistic characteristics. Finally, each object plane is enhanced by the traditional histogram equalization algorithm within the new dynamic range respectively. Because the proposed adaptive segmentation algorithm can automatically segment an image into different object planes and enhance the image according to the statistic characteristic of each object plane, each object components of the image can be well enhanced. Experimental results for different contrast images and the comparisons for some of the previous studies are provided to demonstrate the robustness, visual quality, and effectiveness of the proposed algorithm.