The Design For Image Detection System Of Wheel-rail Relative Displacement Based On Qt

The development of high-speed train technology has brought great convenience to people’s lives; but the locomotive speed and heavy-load are gadually increasing, which can cause interaction movement more intensely and bring about some malfunctions between the wheel and rail trains, such as abrasion、fatigue and derailment. The state of wheel-rail contact is the basic research content in relationship of the wheel-rail, and the change of wheel-rail contact point is the most direct reaction for wheel-rail contact status. Therefore, the measurement of wheel-rail contact point plays an important role in monitor locomotive safety and stability. In order to detect wheel-rail relative displacement, the detection system of digital image based on Qt is designed in this paper.In the process for wheel-rail relative displacement detection, firstly, the method that detection wheel-rail relative lateral displacement based on combination of camera and laser, is proposed in the paper. The camera and laser are located in the bottom of locomotive, and laser coordinate change on the side of track can reflect the wheel-rail relative lateral displacement. After further research, the method that detection wheel-rail relative displacement based on camera is installed on the bogie vertically, is put forward. The camera is installed on the bogie vertically and remains static to bogie, and the wheel-rail relative lateral displacement is measured by rail track coordinate change in the image; simultaneously, laser is installed on the bogie and remains static to camera, which must be shined to track and laser ordinate change can reflect the wheel-rail relative vertical displacement.In the first method, the wheel-rail images have low contrast of light and shade, hence histogram equalization is used to make images clear. Laser point is mapped on the image into small area, for the purpose of obtaining central pixel position in the laser region; threshold segmentation algorithm is utilized to roughly find the most brightness pixel and region growing algorithm is applied to precisely calculated the central pixel position. In the other method, the system has functions of detecting wheel-rail relative lateral displacement、displaying data and storing data under two different conditions of ballast track and unballasted track. Because of terrible detection environment, the track images are made smoothed filter by meanshift algorithm, then, the hough transform is used to extract track edge line and detect wheel-rail relative displacement detection; the means of wheel-rail relative vertical displacement detection is the same as the first method to get ordinate of laser in the image. Finally, wheel-rail relative displacemen can be accurately computed.The experiment results show that the system can acquire wheel-rail relative displacement and has strong adaptability to the detection environment, and it has great important significance for further exploring and evaluating safety mechanics of locomotive.