Retrieval of 2D objects and shape matching using the B-splines representation

Shape retrieval process is composed of two components: shape representation and matching algorithm. In this paper, we propose a new technique for shape similarity. According to this method, the contours are extracted and decomposed into portions of curves at the concavities points. Each portion curve is described by some parametric curve using the B-spline approximation. A cubic B-spline curve is used instead of a higher degree because it has a local control property and is less wiggly. The obtained B-spline curves are then normalized in order to make the method invariant to scale change. This technique uses simple features extracted at high curvature points: Invariant moments and distances from the centroid. Finally the resulting curves are used to compare and to compute similarity between shapes in images database using the L infinity norm. The method has been tested on real images and the experimental results show the performance of the proposed technique.