Direct Non-Symmetry and Anti-Packing Pattern Representation Model of Medical Images

Abstract

With the development of medical treatment, various kinds of medical images become more and more important. Most of them are multi-valued image. Although the hierarchical structural image representation methods have many merits, they put too much emphasis upon the symmetry of segmentation. Therefore, they are not the optimal image representation methods. With the concept of packing problem, a dNAM-based (direct non-symmetry and anti-packing pattern representation model) multi-valued image presentation method is presented in this paper. By describing dNAM, the coding and decoding algorithms of direct NAM method are given. The complexity of the algorithm is analyzed. The theoretical results presented in this paper show that the direct NAM method has not only better compression ability but also has better reconstruction quality. The experiments of four typical medical images prove the conclusion. Some image operation and computation algorithms are also given, they are more effective than the traditional algorithms. The method is a new theoretical research area and valuable for the medical image representation.