Underwater Target Multi-feature Fusion Recognition Technology Based on D-S Evidence Reasoning

摘要

Underwater target recognition is a key technique in military command system. Decision recognition is the last key step. It fully makes several sensor, several kinds of feature and classifier to be combined to form the uniform description to the target according to the optimizing criterion. The paper mainly studies Dempster-Shafer decision theory and its application. Based on the theory, the target feature and Fuzzy Adaptive Resonance Theory (FART) network is combined to form the decision fusion domination. The results show that the decision recognition ratio improves percent two than the single classifier and single feature. It shows that the application of the method is well in target recognition fields.