Surface reconstruction of complex contour lines by means of chain code matching technique

This paper proposes a method to solve the tiling problem in the surface reconstruction. The triangulation method relies on a string matching technique, to find an optimal correspondence between points of previously extracted contours in CT slices. A shape similarity based on a parallelism criterion, is expressed through a cost function and the optimization process is carried out through a dynamic programming algorithm. The surface was then piecemeal built from the set of pairs of matched points. Preliminary results show the method exhibits a good behavior in presence of large variation of the shapes. The time complexity is proportional to M*N (M and N being the number of vertices of each feature respectively in each slice).