Research On Underground Locomotive Crashworthy System Based On Machine Vision

With coal markets remain booming, the underground transportation task of coalmine is much heavier than before.The high-speed locomotives may have transportaccidents in the underground harsh environment.So the research about undergroundlocomotive crashworthy system has an important practical significance to protect theminers’ lifes and guarantee the coal safety production.At present the crashworthy system based on machine vision is mainly aimed at thehighway environment, the researchs about underground coal mine environment aremuch less.The traditional crashworthy systems use infrared, laser or ultrasonic to collectthe data, which have large equipments, high prices and imperfectly effects.In view ofabove questions, this paper brings the machine vision in the underground locomotivecrashworthy system to realize the crashworthy early warning function, by usingmachine visual sensor to detect the obstacle which in the dangerous area ahead oflocomotive.By analyzing the traditional image processing algorithm, considering theunderground factors such as uneven distributing light, paper proposes a kind of imageprocessing algorithm which is suitable for the coal mine underground environment. Thisalgorithm stretchs the gray image to protrude the edge information and reduce theinfluence of uneven light. Through the analysis of traditional edge extraction operatorand the underground rail characteristics, paper proposes a better rail edge extractionalgorithm which combines the gray pixel and gradient pixel together.By designing analgorithm based on edge image to detect the rail inside edge point, paper reduces thecalculation amount of track recognition.According to matlab simulation, results showthat this algorithm can detect the rail and obstacles effectively.Based on the research about the image processing algorithm which is suitable forunderground environment, the paper uses SOPC system to realize the undergroundlocomotive crashworthy function. Compare with the other composite structures, SOPCsystem has less chips, lower prices and less probability of hardware failure.The SOPCsystem uses FPGA logic resource to realize the hardware acceleration methods such asgray pixels real-time processing, edge detection flow design and image parallel storage,which improving the system performance.After debugging, SOPC system can realize the crashworthy function in theunderground environment.The underground locomotive crashworthy system is stable working, achieving the design objective.Based on the result of research, papersummarizes the system and gives a outlook of future development.