A feature level fusion target recognition algorithm based on dynamic fuzzy integral

Fuzzy integral is an effective and convenient fusion algorithm and the key issue is to determine the fuzzy measure, which reflects the degree of importance of each characteristic when it applies to target recognition in feature level. In this paper, after introducing the concept of Choquet fuzzy integral and fuzzy measure, the feature level fusion target recognition converts to the generalized Lebesgue integral of the each feature vector on its degree of importance. Currently fuzzy measure calculation methods only use the static prior knowledge about the training samples. Aimed at a comprehensive recognition process, this paper proposes an adaptive fuzzy integral fusion algorithm, so that the results can be dynamically adaptively corrected in the recognition process.