Research Of Microvascular Recognition Based On Machine Vision

Vessel image segmentation is an important research field in the medical image segmentation which is a main branch of the medical image processing. Drawing blood vessel outline clearly and accurately from the medical images not only can help patients receive the adjunctive therapies but also can provide references of the surgical path demonstration before the surgical operations. Clear and accurate vessel images are essential for rising success rate in surgical operations and allaying patients’painful. With the development of the image collection technology, drawing the blood vessel outline from the contactless observed microvasculature images by using the machine vision equipments becomes a new research subjects. This paper mainly studies the methods of drawing blood vessel outline from the medical images.At present, the vessels images drawing method is mainly applied to the retinal blood vessels images. Although the vessel analysis processing technology of drawing the retinal blood vessels images is consummating day by day, the drawing process is quite complex and the drawing equipment must be the technical fondues camera. The drawing process in the bulbar conjunctiva vessels images is comparatively simple. What’s more, the algorithm running speed in collecting the bulbar conjunctiva vessels images is also comparatively fast. Based on the pathology in the retinal blood vessels is similar to the bulbar conjunctiva vessels, this paper chooses the rampancy web-footed to the observed subject, and draw the contactless observed blood vessels images from the surface of the bulbar conjunctiva by using the multiplication Videocon. In order to compare the merits and drawbacks of the some kinds of methods in vessel drawing algorithm, this paper analyzes the bulbar conjunctiva vessels images with the vision acquirable technology and the digital image treatment technology. Through the above study, this paper offers the new drawing blood vessel outline method which is combining the mathematical morphology and the OTSU from the medical images. The advantage of the new method is that the parallelism of the mathematical morphology and the OTSU minimizes the misclassification object probability. On account of the advantage and the demonstrated algorithm practicability, this paper comparatively improves the immediacy in running algorithm which can be apply to the drawing process in collected images and validate the accuracy in the drawing process. The new method this paper offered can draw the blood vessel outline from the medical images accurately and efficiently which is helpful for guaranteeing to set up the medical treatment detecting system with computer assisted.