An Improved Chan-Vese Model
He Ruiying · 2021 IEEE Asia-Pacific Conference on Image Processing, Electronics and Computers (IPEC) · 2021
Chan-Vese model is a typical geometric active contour model which is able to effectively detect weak edge and discontinuous edge by virtue of global information of image, with favorable noise resistance; however, periodical reinitialization of level set function is required during the process of its curve evolution, which will inevitably increase the time for detection of targets. In this paper, a penalty term is added into Chan-Vese model, and a new geometric active contour model is proposed to overcome the disadvantage of Chan-Vese model for re-initialization of level set function. The experiment shows that the new model will promote the speed of curve evolution, and the initial curve can be defined as any closed shape.