Comparative Study of Segmentation Techniques Used for Optic Disc Segmentation

Fundus image, picture of posterior portion of eye, facilitates non-intrusive diagnosis of various eye diseases. Glaucoma is progressive neurodegenerative ocular disorder, second leading cause of blindness in the world. For diagnosis of Glaucoma using fundus image, cup to disc ratio is the method preferred by researchers. Various methods like Active Contour, K-mean Clustering and Thresholding method are used to segment cup and disk from fundus images. In this paper a comparative study of the above mentioned methods has been presented. These methods were independently tested on two public databases namely DRISHTI-GS, and DRIONS-DB. Based upon the results obtained it was found that active contour technique is best among these methods.