A New Approach to Extracting Boundary of Retina Blood Vessel

Since there are strongly morphologic characteristics and complicated backgrounds in blood vessel images, traditional sharpening algorithms often incur a lot of noises and discontinuities in image boundary. It is necessary to deal with the connection of edges for further processing to obtain the integrity boundary. But by now, the image´s directional characteristics, one of the most useful facts in post processing, are usually neglected and may not be used directly. In order to suppress noise more effectively, a new approach to extracting vascular boundary based on direction filtering is presented in this paper, which is consisted of several algorithms. The method is used in retinal vascular boundary extraction and has achieved good effect.