A vehicle external contour detection and identification system based on multi-sensors

Using laser ranging sensor, radar speedometer and digital camera, a vehicle external contour detection and identification system is designed. According to the data measured from vehicles on the highway, the six features such as vehicle length, height, width, height variance, crest length ratio and vehicle body length ratio are extracted. Then the vehicles are classified to cars, microbuses, trucks, buses, large tracks and large buses automatically by a BP neural network. The experimental results show that this system has better classification effect than existing results, especially could more effectively distinguish the trucks and buses which have similar shape.