Research Of The Detection Algorithm And Design The System On High Speed Railway Fastener Based On DSP

Rail fastener is one of the most important objects to detect in the railway inspection. The train derailment incident may happen if several successive fasteners be lost. In recent years, the more rapid development of high-speed railway technology, the higher requirements are put forward for railway maintenance work in China. Currently, detection of railway inspection still relies on Railway Handcarts and workers in China. However, with the extensive application of computer technology, the railway inspection will be intelligent, integration and real-time detection certainly.The algorithm of the existing fasteners detection has been analyzed in this thesis, finally selected the elastic rail clips as the object to detect. A high-speed efficient new algorithm for detecting missing fasteners has been designed based on the image processing technology. This DSP-based algorithm can be used on the train with the speed up to 300km/h. The research content of this thesis mainly includes discussing on the whole system design, researching on the high efficient fastener detection algorithm, establishing hardware platform of the system and calculating the parameters,selecting the device of each subsystem.This thesis presents the overall design scheme of the system which taking the DSP as the data processing platform. The image captured by the image acquisition system is processed through the algorithm of deletion fasteners detection. Finally the results will be transmitted to the host computer as a later reference for manual inspection.Designing of the image processing algorithm of the fastener detection is the focus of this thesis. An improved Hough line detection algorithm is used in the T-shaped fastener positioning algorithm, and the latest fractional integral filtering algorithm which can ensure well preserved edge during the image de-noising is also applied to the image preprocessing algorithm. In the fasteners detection algorithm, a new algorithm is introduced which is based on image’s local orientation mean to detect the edge of the characteristics enhanced image, and in other way to further strengthen the information of edge image. On this basis, with the method of point by point scanning, this algorithm completes the detection of missing fasteners finally. After the Laboratory verification, this algorithm has the Advantage of fast processing speed and high recognition efficiency.Eventually, after designing the perfect algorithm of fasteners detection, this thesis analyses the actual sample data parameters, completes a data processing system based on DSP platform and selects the hardware equipment. This system achieves the real-time online detection of the fasteners, combining the image processing technology and DSP technology together. Compared with the way of transferring sample image to the host computer by long-distance for processing, this system greatly reduces the noise in the process of transmission, improves the work efficiency and reduces the cost of the system.