An Improved Snake-Based Method for Object Contour Detection

In this paper we present a snake-based method for efficiently detecting contours of objects with boundary concavities. The proposed method is composed of two steps. First, the object´s boundary is detected using the proposed snake model. Second, snake points are optimized by inserting new points and deleting unnecessary points to better describe the object´s boundary. We use the Frenet formula to calculate the binormal vector at snake points and use a regional similarity energy to prevent snake points from converging on foreign edges. Moreover, we use the result to control the direction of movement for snake points near boundary concavities. The proposed algorithm can successfully detect boundary of objects. Experimental results have shown that our algorithm produces more accurate contour detection results than the conventional algorithm.