By comparing the different methods of vehicle ranging, the stereo vision system is selected for obstacle imaging and ranging. In response to low data transmission rate and low real-time performance and low resolution of the traditional ranging system design method, the stereo vision system is designed with integrated multi-core ARM and FPGA SOC. The system adopts image acquisition, distortion correction, target detection and tracking the image of the obstacle in front of the vehicle to compute the obstacle distance information. The delay and the power consumption of the system is measured by experiments, and it indicates that the system is suitable for vehicle-assisted driving and other image processing technologies field.