The current dynamic target recognition methods cannot collect enough feature information in complex images, which leads to the limitation of the application of recognition results. Based on 3D laser point cloud, a moving image dynamic target recognition method is proposed. The 3D scanning system is used to obtain the image features of point cloud, and the image preprocessing is carried out without affecting the effective information acquisition. The ground plane equation is introduced, and the image background point cloud and the identified target point cloud are segmented by Euclidean clustering method. The key points of the identified target are extracted after processing, and the Freeman chain code is used to detect the edge features, so as to complete the dynamic target recognition of the moving image. Experimental results show that the dynamic target recognition method based on three-dimensional laser point cloud has good identification ability, good recognition accuracy and short dynamic target recognition time.