Segmentation of Exudates from Retinal Image

Background/Objectives: Exudates are the bright lesions and major cause for blindness. They are the primary signs of diabetic retinopathy. Due to the damage caused in the blood vessels the exudates are formed in the eye and then they form lesions in the retina. They are hard white or yellow in color with different shapes or sizes. They are to be detected at the early stage. Methods/Statistical Analysis: Exudates are the bright lesions. Optic disc is similar to exudates as they have the same intensity as exudates. The edges were detected using the edge detection techniques after this finally the exudates are detected. Findings: This paper deals with the retinal images to detect exudates. Exudates can be detected easily. In some cases where the intensity is low is also detected asexudates so we need to eliminate these unwanted exudates. Work was performed to detect exudates based different techniques. Improvements/Applications: The earlier mentioned techniques used dilated pupils where the exudates as well as the other retinal features were visible. Non-uniform and low contrast images cannot be used for detection. They are reduced using the preprocessing techniques.