2D Shape Measures for Computer Vision

Abstract

Shape is a critical element of computer vision systems, and can be used in many ways and for many applications. Examples include classification, partitioning, grouping, registration, data mining, and content based image retrieval. A variety of schemes that compute global shape measures, which can be categorized as techniques based on minimum bounding rectangles, other bounding primitives, fitted shape models, geometric moments, and Fourier descriptors are described.