Research on Vehicle Motion Control Strategy Based on Machine Vision

摘要

This paper proposes a machine vision-based detection scheme to complete the automatic advancement and overtaking control of the vehicle. Firstly, the optimal segmentation threshold of the acquired image is obtained by the optimized maximum inter-class variance method, and the threshold is used to binarize the image to obtain the track information, realize the direction control, complete the automatic travel of the vehicle, and then adopt the region growth. The method detects the image of the preceding car in the track, and then obtains the distance between the two cars to complete the detection and control of the distance. Finally, the optimized inter-frame difference method is used to dynamically detect the overtaking behavior and complete the overtaking process. Compared with the traditional way. this method has certain advantages.