Automated Optical Disc Segmentation and Blood Vessel Extraction for Fundus Images Using Ophthalmic Image Processing

Abstract

Diabetic Retinopathy that is characterised by the progressive deterioration in retinal blood vessels is considered the root cause of severe vision loss in diabetic patients. This situation can be reduced upto much extent by regular screening and diagnosis. Precise automatic segmentation of optical disc and automated blood vessel extraction results in effective diagnosis of diabetic retinopathy reducing the chances of vision loss. A considerable progress has been made by various researchers towards automating the ophthalmic image processing via computer aided screening but maintaining the image quality as that of the original fundus image is still a challenge. In this paper, authors have evaluated Optical Disc Segmentation methods based on thresholding, region growing algorithm and mathematical morphology for effective removal of optical disc to facilitate blood vessel extraction. A new blood vessel extraction technique using Mathematical Morphology and Fuzzy Algorithm is proposed for precise blood vessel extraction. Two open access standard fundus image databases, DRIVE and STARE were exploited for performance evaluation of the proposed approach. This approach is effective in identifying optical disc to extract the blood vessels near the optical disc area which plays an important role in early diagnosis of diabetic retinopathy.