Structural Object Matching

The skeleton and its associated medial axis give a very compact representation of objects, even in the case of complex shapes and topologies. They are powerful shape descriptors, bridging the gap between low-level and high-level object representations. Surprisingly, skeletons have been used in a relatively small number of applications, and almost every time for very precise and concrete tasks. This work focuses on the problem of object registration using the objects’ medial axis. We first build adequate attributed relational graphs to organize in a structured way informations about object shape and topology contained in the medial axis. Using a graph matching algorithm then allows to solve the correspondence problem between the graphs. From these correspondences between similar parts of the objects, we infer a set of matched points and finally estimate the transformation between the two objects through a robust matching algorithm. Synthetic results are presented.