A Novel Technique for Fundus Image Contrast Enhancement

Digital fundus Image analysis plays a vital role in computer aided diagnosis of several disorders. Image acquired with fundus camera often have low grey level contrast and dynamic range .We present a new method for fundus image contrast enhancement using Discrete Wavelet Transform (DWT) and Singular Value Decomposition(SVD).The performance of this technique is better than conventional and state of the art-techniques. With the proposed method the given Fundus Image is decomposed into four frequency sub band images and Singular Value Decomposition applied on Low-Low subband Image, which determines the intensity information. Finally Image reconstructed using modified Low-Low subband coefficients and three high frequency sub band coefficients. The qualitative and quantitative performance of proposed technique is compared with those of conventional techniques such as General Histogram Equalization (GHE), Local Adaptive Histogram Equalization (AHE), Contrast Limited Adaptive Histogram Equalization (CLHE), Brightness Fuzzy Preserving Histogram Equalization (BPFHE) and Singular Value Equalization (SVE). The proposed method shows better results than other contrast Enhancement Techniques on low contrast and dynamic range fundus Images, which leads to give helpful for retinal segmentation.