No abstract

No abstract

No abstract

The WARP system defines a dissimilarity measure between shapes described by their contours which is based on Dynamic Time Warping of Fourier Descriptors based signatures. These signatures are invariant to translation, scaling, rotation, and selection of the starting point. However, identical shapes present ambiguous signatures and similar shapes may yield significantly different signatures. Differences affect rotation and starting-point of the signatures, which may lead to poor performance in classification and shape retrieval tasks. We propose a different signature method to provide true rotation invariance and a Cyclic Dynamic Time Warping dissimilarity measure to achieve true startingpoint invariance in shape comparisons. This work has been supported by the Spanish Ministerio de Ciencia y Tecnología and FEDER under grant TIC2002-02684.

Abstract. The cyclic edit distance between two strings A and B of lengths m and n is the minimum edit distance between A and every cyclic shift of B. This can be applied, for instance, in classification tasks where strings represent the contour of objects. Bunke and Bühler proposed an algorithm that approximates the cyclic edit distance in time O(mn). In this paper we show how to apply a technique for ranking the K shortest paths to an edit graph underlying the Bunke and Bühler algorithm to obtain the exact solution. This technique, combined with pruning rules, leads to an efficient and exact procedure for nearest-neighbour classification based on cyclic edit distances. Experimental results show that the proposed method can be used to classify handwritten digits using the exact cyclic edit distance with only a small increase in computing time with respect to the original Bunke and Bühler algorithm.