Abstract

Computer-aided medical diagnostic system design is an emerging to assists medical practitioners for providing\nquick and accurate diagnosis and prognosis of pathology In this Database, we study new methods to analyze digital fundus\nimages. In particular, we concentrate on the development of the algorithmic components of an automatic screening system for\ndiabetic retinopathy. The techniques developed can be categorized broadly in quality assessment and improvement, lesion\nsegmentation. firstly, we present a fast algorithm to numerically calculate approximately the quality of a single image by\nemploying vasculature and color-based features, that is we convert given image to respective Red, Green and Blue\ncomponents (RGB) additionally, we show how it is possible to increase the image quality and remove reflection artifacts by\nmerging information gathered in multiple fundus images (which are captured by changing the stare point of the patient). It is\nseen in the images that the blood vessels are best visible in Green component of image. Then the green image is converted\ninto grayscale image and enhanced. Using Gabor Filter is used for edge detection as it is seen that the Frequency and\norientation representations of Gabor filters are similar to Human Visual system. Under certain conditions the phase of the\nresponse of Gabor filters is approximately linear and these properties serves the best use in detection of edges and then we\ncan easily identify the swelling of the macular.