Glaucoma Screening Through Level Set for Optic Disc Segmentation and Textural Features for Classification

Glaucoma is considered as world's second biggest health emergency of this century. Glaucoma is an irremediable chronic ocular disease. In Glaucoma, Optic Nerve Head (ONH) gets damaged due to loss of ganglion cells. This damage of ONH causes cupping of nerve head and optic cup size increases with respect to the size of optic disc. These changes in ONH can be detected using various image processing methods which can be applied on retinal images acquired using Optical Coherence Tomograph, Retinal fundus camera and Heidelberg Retina Tomograph. This paper proposes an approach to detect Glaucoma automatically from retinal fundus images. A level set based approach for segmentation of optic disc (OD) is used and further analysis of segmented OD is done by textural information. In this approach, image features are obtained by the use of Local Binary Pattern (LBP) and these image features were classified using Support Vector Machine(SVM) classifier.