A Vehicle Detection System Based On Visual Image Of The Cable Fault

In order to ensure the safety of railway operation, the reliable operation of GSM-R leakage cable communication in Mountainous Tunnel is needed in the course of railway operation. Because of the low efficiency of detection, the need to open the window and other issues at the present stage, so the paper design a vehicle detection system based on visual inspection and image processing, under the condition of not disturbing the railway transportation order, the overall image of the railway leakage is acquired, the image is compressed and transmitted, at last, the fault detection and identification of the leakage of the cable is carried out, in time to prevent the leakage of the loose, loose or broken, etc.In this paper, we design a reasonable technical scheme for high-speed running vehicle image acquisition system, such as the high speed of acquisition, the weak light, the large amount of information and so on. First of all, this paper is based on the study of image compression algorithm, and the image compression is carried out by using JPEG image compression algorithm, and FPGA is selected to achieve the image compression, which is of great capacity and processing speed. On the basis of analyzing the characteristics of the fault of the cable, the cable is located by the Snake feature detection method, Haar method to extract characteristics of fixture, and SVM are combined to realize leakage cable clamps of the positioning. Based on texture feature LBP extraction method combined with SVM method were used for fault recognition. Aiming at the problem that the system needs to be solved in the system of the fault detection system based on visual image, the key technologies of image acquisition, effective field of view, illumination and automatic acquisition are studied. According to the specific requirements of image acquisition, image processing and the special points of railway tunnel, the overall structure and modules of the hardware system are designed; software system is designed and image acquisition and compression storage, image fault identification and other software modules are developed respectively.Finally, the test system of the fault detection system of the vehicle on the high speed train is carried out. The experimental results show that the detection system can achieve the detection of the cable fault, which has important significance for the safety operation of the railway wireless communication system, and promote the development of railway productivity.