Aiming at the requirement of automatic target recognition, a method of laser and infrared fusion target recognition is proposed. Firstly, wavelet moment and projection contour feature are extracted respectively from laser and infrared datas respectively. Secondly, two kinds of single source features are combined, and then the features are reduced. Considering the combination of multiple reductions will produce complementary information, three different views of the reduction algorithm are adopted to construct the diversity combination classifier. Then the fusion recognition method of laser and infrared fusion is presented. Finally, the effectiveness of the proposed method is verified by the experiment of laser and infrared simulation data.