An Improved Object Detection and Contour Tracking Algorithm Based on Local Curvature

Using the classical snake algorithm it is difficult to detect the contour of an object with complex concavities. Whereas the GVF (Gradient Vector Flow) method successfully detects the concavity of a contour, but consumes lots of time to compute the energy map. In this paper, we propose a fast snake algorithm to reduce computation time and to improve the performance of detecting and tracking the contour. In order to represent the object's contour accurately, a snake point inserting and deleting strategy is also proposed. Simulation results from a sequence of images show that our method performs well in detecting and tracking the object's contour.