Retinal vessel extraction using novel multi-scale hessian-based enhancement and radial gradient symmetry transform

In this paper, we propose a new method to extract the blood vessels in retinal images. First, we perform a novel multi-scale hessian-based filter to detect the maximum response of vessel likeness function, which is used to enhance the blood vessels of grey retinal images. After that, a new radial gradient symmetry transformation is adopted to improve the detection of vessel structures and suppress the non-vessel structures. Finally, the central lines of the detected vessels are extracted to give a visual comparative result with manually segmented vessels. We test the proposed approach on the publicly available databases-DRIVE. The experimental results show the effectiveness of our method.