Morphological contour coding using structuring functions optimized by genetic algorithms

Shape representation is an important image analysis task which can be used for contour coding and feature extraction. The morphological skeleton is a geometrical shape description by means of maximal inscribed structuring elements. The form of the structuring element is usually chosen a priori, and we show how genetic algorithms can be used for an automatic optimization of an arbitrary shaped structuring element. It permits improved progressive contour transmission and the extraction of shape features.