Study And Realization On The Train Fixed Hook Detection System Based On ARM9

Dumper is a high efficiency turn unloading equipment, mainly used in the occasionwhere have rich mineral resources, like, port, power plants, mines and so on. With theimprovement of the dumper automation degree and the unity of railway vehicle of opentop wagon the unity, appeared not pick hook dumper system, but the premise of thedumper system security operations is the train car with rotation hook, and carriages’connection in correct order. But now mostly used artificial cognition method, this wouldin a certain extent influence the dumper overturned efficiencyThis paper designed the train fixed hook detection system based on ARM embedded,in line with three car dumper not pick the hook when Unloading operations inQinhuangdao Port. The main contents are outlined below:Firstly, this paper analyses the train C80 hook status. The Coal trains C80 marks theproperty of the hooks according to different colors of the carriages in Qinhuangdao Port.Takes the fixed hook detection converted into image color discrimination issue,according to color of the train carriages.Secondly, design image recognition algorithm. This paper adopts to use FCMalgorithm, which is improved on CIE Lab space and ant colony algorithm to realize colorclustering. This method uses the global and robustness advantages of ant colonyalgorithm, combines with CIE Lab space and sets clustering center through A and Bcomponent of CIE Lab space, which is to be initial center of FCM and can overcomeFCM local optimal faults. Then do the simulation test, confirm that this algorithm has agood accuracy.Finally, this paper does hardware design by S3C2440A as the main control chip,solve the problem of the train hook detection, which were based on ARM core andCMOS camera, transplant software for the bottom of the Linux operation system toimage acquisition transmission, including image processing, the train hook identify andso on. Do debug in the lab, test and verify feasibility of the system. The result indicatesthat the train fixed hook detection system designed by this paper can get relatively accurate identification results and reach the expected effect.