A target recognition method based on feature level data fusion

To solve the problem of multi-feature fusion target recognition, a feature level fusion method using a multi-classifier is developed. Firstly, the combination of sensors used in target recognition and the model structure of feature level data fusion is discussed. Secondly, a feature level fusion method based on a multi-classifier is presented. In this method, fuzzy logic systems with different expert knowledge are used as classifiers, a parallel structure used as a combination model, and the D-S inference used as the combining method of different classifiers. The simulation result showed that this method can fuse different kinds of features to classify targets effectively.