A-GLG에서 발생하는 대조비 개선 과도 현상의 감소 방법

Abstract

In this paper, we propose an efficient method of reducing excessive contrast enhancement by A-GLG(Adaptive Gray- Level Grouping). A-GLG, a contrast enhancement algorithm, segments an imput image into subimages and performs grouping from a histogram of each subimage to calculate a transformation function. However, A-GLG performs excessive contrast enhancement, if the number of groups in the initial subimage is too small. Besides, the calculation of quality evaluation for contrast enhancement at each grouping of each subimage is time-consuming. The proposed method applies the identity transformation function to overcome the excessive contrast enhancement, by measuring the number of initial groups of each subimage. To reduce the time complexity, the proposed method reduces the number of quality evaluations for contrast enhancement.