Computer Intelligent Image Recognition Algorithm
Gang Wang · Jisuanji gongcheng · 2004
In complex environment, intelligent computer vision system can meet the needs of accurate recognition object. By combining Dempster-Shafer evidence reasoning with invariant moment theory, this paper presents intelligent image recognition algorithm, and the algorithm is applied to two dimension airplane image recognition. Firstly, a basic belief assignment function is constructed, and a classifying rule is built. Furthermore, in order to solve the problem which Dempster combination rule is used only in case of high belief degree and low conflict, the weighed distribution conflict and the absorptive are used to solve evidence conflict. The result of computer airplane image recognition simulation indicates that absorptive has the highest recognition efficiency in the three combination rules.