Analysis of Combination Algorithms for Denoising and Contrast Enhancement Images

Abstract

Reducing noise and increasing image contrast is part of the purpose of enhancing image quality; instead, it will impact change the diversity of information in the image based on the Shannon entropy value. Decrease quality caused by noise salt and pepper in this research or abnormal contrast in the image causes objects in the image to become unclear. Low contrast has a major impact on image quality, including noise reduction processes affecting image information so that the quality of the reduced image becomes something to consider for large noise. Iterative Denoising and Backward Projections with CNN (IDBP-CNN) and Different Applied Median Filter (DAMF) is a good solution for denoising a large percentage of noise with good quality results image. In other research for contrast enhancement, Triangular Fuzzy Membership-Contrast Limited Adaptive Histogram Equalization (TFM-CLAHE) and Adaptive Fuzzy Contrast Enhancement Algorithm with Details Preserving (AFCEDP) is claimed to a good solution to solve low contrast of the image. Therefore, this study is to find the best combination of denoising and contrast enhancement to get good image results with step denoising followed by contrast enhancement. Based on the experimental testing is got the best combination is the DAMF + AFCEDP algorithm with an average of PSNR 35dB and an average difference Shannon entropy of 0.0130.