Contrast Enhancement of Portal Images with Adaptive Histogram Clip

Authors: Georgi GLUHCHEV

Abstract: A novel approach is described aimed at the contrast enhancement of portal images. To avoid the edge blur caused by the histogram equalization techniques achieving a good contrast, at the same time, a particular clip level is assigned to the contextual regions covering the image. For this, regions are classified into clusters B, F and E referring to the background, field and field ege, revpectively. The clustering procedure is based on the mean value m and standard deviation \sigma of the region grey levels. Small clipping levels in the background and edge areas preserve the noise and edges while large values inside the field lead to a significant contrast enhancement. The experiments with portal images have shown the good performance of the approach.

Keywords: contrast enhancement, cluster, adaptive clip, portal image.