An Efficient Approach of Optic Disc Normalization and Segmentation for Glaucoma Detection

Abstract

Glaucoma is considered as one of the major eye disease which will lead to vision loss if it is not diagnosed at a right time. Hence it is required to recognize the stage of the disease as early as possible. The earlier methods called Intra ocular pressure(IOP) and Visual Field Test have a disadvantage of requirement of special equipment which will be available in only specialized hospitals and provide low accuracy. In this paper effective method called Sparse Dissimilarity Constrained Coding (SDC) have been used where it considers optic disc and cup called cup to disc ratio (CDR). In this approach the optic disc is localized and segmented which is followed by cup segmentation. From which the area of optic disc and optic cup is obtained. The method gives accurate CDR results and it is well suited for more population. From the obtained ratio the stage of the disease can be well predicted and suitable treatment required can be suggested. The retinal fundus images that are used for the method will be easily available in almost all the hospitals and medical centers for comparing the result with the reference CDR ratios. The method provides efficient and reliable result compared to the manual method. Hence the proposed method is an effective approach for glaucoma detection.