PROBABILISTIC IMAGE MODEL AND ITS INFORMATION-THEORETIC PROPERTIES

被引:0
|
作者
ZHANG, YQ [1 ]
LOEW, MH [1 ]
PICKHOLTZ, RL [1 ]
机构
[1] GEORGE WASHINGTON UNIV,DEPT ELECT ENGN & COMP SCI,WASHINGTON,DC 20052
关键词
D O I
暂无
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
In this correspondence, we propose a method to construct a class of correlated non-Gaussian source from a specified first-order distribution and correlation structure. The obtained source model is based on a nonlinear transformation from a correlated Gaussian source, and forms a special class of the Gibbs distribution. Its entropy properties and rate-distortion performance are examined. An example is also given to illustrate the construction process and the characteristics of the source model.
引用
收藏
页码:508 / 510
页数:3
相关论文
共 50 条