Performance Evaluation of Joint Filtering and Histogram Equalization Techniques for Retinal Fundus Image Enhancement

Abstract

Fundus retinal images regularly experience the ill effects of non-uniform light, poor contrast and noise. This demands implementation of suitable pre-processing techniques for enhancing the image quality. Pre-processing technique includes methods of noise removal as well as contrast enhancement. Fundus images often altered by impulse noise. Meanwhile, quality of retinal fundus image suffers due to nonuniform illumination and low contrast due to anatomy of eye fundus. This paper presents the performance evaluation of joint noise filtering and histogram equalization techniques for the overall quality enhancement of retinal fundus image. In the first stage of the work, two emerging noise-filtering methods namely, NAFSM filter and block-matching and 3D filtering (BM3D) techniques are implemented for de-noising. In second stage, histogram equalization (HE), adaptive histogram equalization (ADHE), contrast limited adaptive histogram equalization (CLAHE), exposure based sub-image histogram equalization (ESIHE) techniques are investigated one by one for obtaining the best suited combination of the filtering and HE methods for the improvement of retinal fundus picture. Performance evaluation of these techniques have been carried out by determining the performance parameters such as histogram, SNR, entropy, absolute mean brightness error and peak signal-to-noise ratio (PSNR) of processed retinal fundus images.