Retinal Vascular Image Segmentation And Its Application Research

With the rapid development of image processing technology, people can better prognosis for the treatment of certain diseases and the timely recovery of control, so it maks the treatment be greatly improved. Computer image system, many organizations can fundus quantitative measurement and detection, between normal and abnormal to make a clear identification of the system gradually to standardization.This article focuses on the segmentation of retinal blood vessels, the first image on the retina in different directions on a series of operations to carry out filtering operations, these operations include morphology dilation, erosion, re-opened closed operations and reconstruction operations. After these pretreatment can be greatly enhanced the contrast of blood vessels and the background, effectively reducing the retinal image in some of the noise, thus improving the accuracy of extraction of retinal blood vessels. Late registration for retinal and connection, the main blood vessel network is to use the interrection of the centerline and the basis of similarity with vascular network registration method, the image of the connection basically completed the work.In the theoretical analysis and research based on the realization of the segmentation of retinal blood vessels, the basic completion of the automatic registration between the image and link. In particular, achieved good segmentation results, which for the subsequent image registration and image mosaic ,provides a good basis for the connection.