Statistical morphological skeleton for representing and coding noisy shapes

A new shape descriptor obtained by skeletonisation of noisy binary images is presented. Skeleton extraction is performed by using an algorithm based on a new class of parametrised binary morphological operators, taking into account statistical aspects. Parameters are adaptively selected during the successive iterations of the skeletonisation operation to regulate the characteristics of the shape descriptor. A probabilistic interpretation of the scheduling strategy used for parameters is proposed by analogy to stochastic optimisation techniques. Skeletonisation results on patterns extracted by a change-detection method in a visual-based surveillance application are reported. Results show the greater robustness of the proposed method as compared with other morphological approaches.