Detection of Glaucoma by Segmenting Retinal Images and evaluating the DDLS stage

Abstract

Automatic analysis of retina image is important screening tool for early detection of eye diseases. This paper proposes an image processing technique for the detection of stage of glaucoma. The methodology is used to segment optic disc and cup. The method starts with preprocessing of digital fundus image to extract region of interest surrounding optic disc. Hough transform is used to define approximate initial boundary of the optic disc which is then deformed to find out exact optic disc boundary by active contour model. This technique gives effective results to determine rim to disc ratio for detection of glaucoma and thus to detect its DDLS i.e. stage of glaucoma.