Review of shape coding techniques

Abstract

Shape is a concept which is widely understood yet difficult to define. This review reports the techniques used to code shape. The techniques may be grouped in various ways with the most significant distinction being between space domain and scalar transform approaches. Scalar transform approaches use linear summations derived from shape whereas space domain methods consider the structural and relational properties of features. The methods are also split into the categories of internal and external descriptors which distinguish between those techniques which represent the boundary of an object and those which code the area within it. Other properties of the methods considered are presented such as whether the schemes allow reconstruction of shape from the descriptor, their ability to recognize an object from a partial description, computational complexity and their limitations.