Image Processing during the Pursuit of Beacon Lamp by Smart Cars

To improve the beacon positioning technology in automation system, an image-based beacon positioning method is proposed and applied to smart car competition. The real environment is considered and the median filtering algorithm, the dynamic threshold algorithm, the exchange frame discarding algorithm, the improved connected area algorithm and the improved distortion reduction algorithm are used to process images captured by CCD and CMOS cameras. The error rate of this method is analysed. The results of the experiment and the application of the method in competition show that this method can achieve the goal of searching beacon lights for smart cars at high speed. The error rates of 1 000 tests under strong light(about 200 lx) and weak light(about 75 lx)environments are 1.4% and 3.2% respectively. The research results are of great significance for solving the problem of accurate beacon positioning.