Retinal Fundus Image Refinement with Contrast Limited Adaptive Histogram Equalization, Noise Filtration and Intensity Adjustment

Retinal Fundus disease is an explosively increasing research topic today for several reasons. The problem with images is the quality that makes it difficult for health care providers to interpret the condition accurately. In this study, a method has been proposed to improve the quality of retinal fundus to solve this problem. Test data for the research comes from a set of STARE datasets with different procedures, which include combined contrast-limiting histogram equalization, noise filtration, and intensity adjustment. In this study, we used the Mean Square Value (MSE) and the Peak Signal To Noise Ratio (PSNR) as parameters to compare the image quality of retinal fundus of each process. The experimental results of this study justify that these procedures can be used to improve image quality and convert low-quality images to a high-quality image.