Conversion Database of the Shapes Into XML Data for Shape Matching

We present a new approach to the matching of 2D shapes using XML language and dynamic programming. Given a 2D shape, we extract its contour and which is represented by set of points. The contour is divided into curves using corner detection. After, each curve is described by local and global features; these features are coded in a string of symbols and stored in a XML file. Finally, using the dynamic programming, we find the optimal alignment between sequences of symbols. Results are presented and compared with existing methods using MATLAB for KIMIA-25 database and MPEG7 databases.