Abstract: By analyzing the license characteristic and the existing recognition technologies，this paper presents a novel character recognition algorithm based on multi-classifier，an integration of template matching and support vector machine (SVM). In the proposed method, the original problem is firstly question firstly coarsely classified into a two-classes problem by using template matching, and then are accurately classified by means of SVM. In the procedure of template matching, the idea of intensity normalization is proposed in this paper. On the Virtual Studio platform，the testing results show that the proposed algorithm achieves better adaptability and higher recognition accuracy.