Image contrast enhancement for preserving entropy and image visual features

Histogram equalization is essential for low-contrast enhancement in image processing. Several methods have been proposed; however, one of the most critical problems encountered by existing methods is their ability to preserve information in the enhanced image as the original. This research proposes an image enhancement method based on a histogram equalization approach that preserves the entropy and fine details similar to those of the original image. This is achieved through proposed probability density functions (PDFs) that preserve the small gray values of the usual PDF. The method consists of several steps. First, occurrences and clipped histograms are extracted according to the proposed thresholding. Then, they are equalized and used by a proposed transferring function to calculate the new pixel values in the enhanced image. The proposed method is compared with widely used methods such as Clahe, CS, HE, and GTSHE. Experiments using benchmark datasets and entropy, contrast, PSNR, and SSIM measurements are conducted to evaluate the performance. The results show that the proposed method is the only one that preserves the entropy of the enhanced image of the original image. In addition, it is efficient and reliable in enhancing image quality. This method preserves fine details and improves image quality, supporting computer vision and pattern recognition fields