Quadrants Histogram Equalization with a Clipping Limit for image enhancement

The paper proposes a novel image enhancement method based on histogram equalization called Quadrants Histogram Equalization with a Clipping Limit (QHECL). The proposed method consists of four steps: (i) The first step is to compute the median value of the input image, which is used to divide the input histogram into two sub histograms. (ii) We calculate the average brightness value of each sub histogram and use the average brightness value to segment each sub histogram respectively. Thus, we can get four sub histograms. (iii) The average of the occupied intensity is computed and viewed as a plateau limit to clip each sub histogram. (iv) Four sub images are enhanced by the tranditional histogram equalization method independently and then are merged into one output ehanced image. The experimental results show that QHECL is better than other contrast enhancement methods according to image subjective evaluation and various image objective evalution measures, i.e. average information content (Entropy), PSNR, absolute mean brightness error (AMBE).