Adaptive morphological operators, fast algorithms and their applications

A general form of morphological operators is developed. The structuring elements of the operators adapt their shapes according to the local features of the processed image and can be any shape formed by connecting a given number of predetermined basic elements. The basic element can be a single pixel or any other connected small shape. It is shown that while sharing the basic properties of the conventional morphological operators, the operators also have some distinguished properties. Those properties are related to intuitive descriptions of the geometric performances of the operators and the development of fast algorithms for the implementation of the operators. The efficiency of the operators in image processing is discussed, and two application examples are given.<>