Automatic glaucoma screening using superpixel segmentation

Glaucoma is the second leading cause of loss of vision in the world. Currently, there is no effective method for simple, accurate and low cost glaucoma detection or screening. Automatic optic nerve head assessment from 2D retinal fundus images is promising for low cost glaucoma screening. This paper presents efficient method for cup to disc ratio (CDR) assessment using 2D retinal fundus images. The disc segmentation is carried out using active shape model (ASM), elliptical hough transform (EHT) and superpixel segmentation techniques. Among these three methods, one best output depending upon self assessment confidence score is considered for further processing. Cup to disc (CDR) ratio is calculated for the assessment of glaucoma. Based on performance of each method superpixel method performs best and more reliable.