Robust retinal blood vessel segmentation to detect diabetic retinopathy

As doctors are facing problems for obtaining the retinal segmented image of patient eye with in less time, effort and with more efficiency, there is a need of reliable automated method for retinal blood vessel segmentation in computer-aided diagnosis. Moreover, the retinal vascular tree is unique for each individual and can be used for biometric identification for high security applications. It is commonly accepted by the medical experts that the automatic segmentation of retinal vessels is the first step for the development of a computer-assisted diagnostic system for ophthalmic problems. So, it is proposed to develop a methodology for the doctors to segment the blood vessels of the patient digital retinal scan within few minutes using the proposed image processing technique on patient digital scan which has a 97% accuracy in comparison to manual images generated or segmented by ophthalmologists. In this paper, it is proposed to apply various segmentation techniques and filters on different images available in benchmark dataset, DRIVE.