A MATLAB implementation of Chan-Vese model for image segmentation, using idea proposed by Chan and Vese. For detailed information, please refer to the paper "Active Contours Without Edges", or my blog.
- The test images are placed in "images" folder;
- You can load the existing initial contours from "initial contours" folder or select a polygonal region by hand;
- Also, you can record the evolution process of active contours in a GIF file;
- Segmentation results will be saved in "results" folder;
- Run "demo_CV.m" directly to get segmentation results.
NOTE: To use it as academic purpose, please cite the following papers:
 Active Contours Without Edges, Tony F. Chan and Luminita A. Vese. IEEE Trans on Image Processing, 2001.
 Level Set Evolution Without Re-initialization: A New Variational Formulation, Chunming Li, Chenyang Xu, Changfeng Gui, and Martin D. Fox. IEEE CVPR, 2005.