In order to solve the problem of vehicle information acquiring, system and method for vehicle feature recognition are designed based on laser scanning 3D imaging. 3D point clouds are obtained via a 3D scanning system based on a 2D laser scanner carried on an electric turntable. Method for system self-demarcation is developed based on point clouds, which can automatically calculate the placement parameters of the system. Moreover, vehicle feature recognition method is presented, in which the point clouds are segmented into different classification, and characteristic parameters are calculated. System control and data process software is designed and realized based on QT, in which system control, data collection and process, data exchange and control communication are realized. Experiments are presented in scene of car loader, experimental results show that the system can reach the measure accuracy of 5 cm and has a short time cast less than 40 s, which can be applied in vehicle feature recognition for car loader.