Automated retinal blood vessels extraction using Optimized Gabor filter

摘要

This paper presents a new automatic method to extract the retinal blood vessel under various normal or abnormal conditions. Extraction of blood vessels in retina is helpful for ophthalmologists to screen larger populations. The commonly used Gabor filter methods often produce false positive detections of blood vessels. We proposed a method for segmentation of blood vessels using Optimized Gabor filter as they tuned to the specific frequency and orientation. We have employed robust performance analysis to evaluate the accuracy. Our proposed Optimized Gabor filter demonstrated higher true positive rate and reduce false detection than existing various techniques in vessels extraction.