Snake-Based Objects Tracking in Stereo Sequences with the Optimization of the Number of Snake Points

Optimizing the number and placement of snake points is very crucial to the performance of the snake algorithm especially when used for object contour tracking. In this paper we address the problem by managing the insertion of new points and deletion of unnecessary points to better describe and track the object's boundary. In particular, our method uses more points in highly curved parts of the contour, and fewer points in less curved parts. The proposed algorithm can successfully define the contour of the object, and can track the contour in complex images. Performance of the proposed algorithm has been verified with a set of experiments.