A Novel Fusion Filtering Algorithm Based on D-S Theory  

作　　者：WANG Xuanming ZHANG Xiaolin JING Xiaojun 

机构地区：[1]Beihang University, Beijing 100191, China [2]Beijing University of Posts and Telecommunications, Beijing 100876, China

出　　处：《Chinese Journal of Electronics》2011年第3期579-582,共4页电子学报（英文版）

基　　金：This work is supported by the National Natural Science Foundation of China （No.60872148）.

摘　　要：A novel fusion filtering algorithm is pro- posed to resolve the problems of the hybrid filtering algorithm based on advanced D-S theory of evidence. The hybrid filtering algorithm judge image edges. However, for the image pixel, the hybrid filtering algorithm adopts the single method that is linear or non-linear, and it is not a real fusion filtering algorithm. Accordingly, the proposed algorithm based on fusion theory judges image edges, but also makes fusion filtering, which improves the performance of filtering. The experimental results indicate that the proposed algorithm is effective.

关 键 词：Information fusion D-S theory of evidence Minimum total error probability criterion NeymanPearson criterion Fusion filtering. 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置] TN911.7[自动化与计算机技术—控制科学与工程]