Performance analysis of matched filter techniques for automated detection of blood vessels in retinal images

The paper addresses issues in the development of an automated system for the analysis of retinal angiographic images, focusing on the segmentation of the blood vessels. The performance of three different template matching algorithms are analyzed in respect of the detection of blood vessels in retinal images for both gray-level and color images. The Gaussian matched filter (GMF) and binary matched filter (BMF) are used to detect the edges of blood vessels and capillaries of gray-level images and also to reduce the noise. The Kirsch template matched filter (KMF) is used for the same purpose in color images. The results obtained provide the complete vessel map, thereby making diagnosis easier for the ophthalmologist.