A Review and Survey Paper on Image Processing

A Review and Survey Paper on Image Processing

Vivek Verma, Anant Dewangan, Himanshu

Department of Information Technology

SSIPMT, Raipur

Abstract — Low- light images are not conducive to mortal observation and computer vision algorithms due to their low visibility. Although numerous image improvement ways have been proposed to break this problem, being styles inescapably introduce discrepancy under- and over-enhancement. In this paper, we propose an exposure emulsion frame and an improvement algorithm to give an accurate discrepancy improvement. Specifically, we first design the weight matrix for image emulsion using illumination estimation ways. In addition, we introduce our camera response model to synthesize multi- exposure images. Next, we find the stylish exposure rate so that the synthetic image is well- exposed in the regions where the original image under- exposed. Eventually, the input image and the synthetic image are fused according to the weight matrix to gain the improvement result. Trials show that our system can gain results with lower discrepancy and lightness deformation compared to that of several state- of- the- art styles.

Keywords — Contrast enhancement, equalization, normal distribution, histogram partition