Comparison of Higher Order Statistical Method for Low Contrast Images

Abstract

Contrast enhancement and brightness preservation are widely used in different application of image processing. Histogram equalization (HE) is a common methods used for contrast enhancement as it is easy to implement for any digital images. But this method is not suitable for electronics application because it introduce unnecessary visual artifacts in the output images. To overcome the visual artifacts several segmentation methods are introduced for contrast enhancement. The two segmentation based on image enhancement algorithms, called Root-Mean-Skewness Bi-histogram Equalization (RMSKBHE)[8] and DUO-Histogram Equalization(DUO-HE)[9] gives the general idea or shape of the histogram of images. In this paper both the RMSKBHE and DUO-HE algorithm are proposed for better contrast enhancement of asymmetric input image and to find better contrast enhancement and brightness preserving for output image. The comparison of RMSKBHE and DUO-HE algorithms have been carried out using different quality measurement technique i.e. peak signal to noise ratio (PSNR) and absolute modified mean brightness error (AMMBE). Visual perception and statistical values for the DUO-HE algorithm is found out better than the RMSKBHE algorithm.