Image Detecting Algorithm Research For The Wear Of Pantograph Slide

Basing on large investigation of the detecting technologies for pantograph slide home and abroad at present, the paper adopts the optics image detecting and processing technology to inspect the wear of the pantograph slide. It is a non-contacting online inspecting way automatically. And it has the advantages such as high efficiency, high detecting precision, no need of stopping. The design and unceasing improvement of the algorithm based on image processing is the key for increasing the precision of the slide abrasion.The system adopts CCD camera to collect the original images of the front and back slide of the pantograph. And then the images are transmitted to the main detection PC and stored there. Through post-processing of the images, the system can inspect the slide abrasion automatically and output the data report and abrasion curve.The image processing algorithm of slide abrasion detection is designed in this paper. It includes image preprocessing, edge detecting and linking, contact wire and slide edge location and etc. And the core and emphasis part is the edge detecting and location algorithm which influences the final measuring precision. Therefore it is the key part of the algorithm design. During the whole designing processing, the paper firstly compares the advantages and disadvantages of several edge detecting and linking algorithm and then validates the processing effect of them by programming the code. The result shows that processing effect of the algorithm of canny edge detecting combined with morphology edge linking is much better. And the location algorithm is realized by Hough transform and edge searching. Finally, the curve of slide thickness remained is obtained by calibration which can convert the pixel coordinate to physical coordinate.As the current image processing algorithm has some limitation for different kind of slide pictures such as aluminous edge pictures, the pictures of LV2600 type pantograph and pictures on the condition of rain and snow. Mostly, they are impossible to be processed and sometimes failed. So some improvement is made afterwards and the testing result shows that it can solve most of the problem above. Furthermore, some improvement of the hardware aspect is also done in the system. Firstly it chooses another industrial strobe which has better performance as the light source of the system. Then for some special model of the pantograph, it changes the shooting way as looking up. Both of them make great improvement of pictures' quality and reduce the difficulty of algorithm design.The result of experiment and application on the spot shows that both the improvement of algorithm and system hardware made the detecting precision increase greatly. And the effect is quite obvious. And the mproving of algorithm made the accurate processing ratio increase greatly. So the paper definitely has some reference value not only for the image inspecting ways of slide wear but also for the design and realization of edge detecting and location algorithm.