Key Technology Research Of Track Static Detection Based On Vision Measurement

With the development of railway moves towards high-speed and overloading, better track regularity of train operation has become increasingly demanded. Track static detection is one of the key technology which guarantee track regularity, and some shortcomings still exist in China’s track static detection equipments. Combined with the advantages of vision measurement, a Track static detection equipment based on vision technology is presented in the paper. The thesis has completed the following works:(1) Combined with the original structure design of T type track measuring instrument, redesigned the mechanism and selected the related hardware, using linear structured light and industrial cameras as the monocular vision measurement system.(2) Analyzed the detection principle and characteristics of original track measuring instruments, a method of track static geometric parameter measurement based on vision measurement technology is presented by using simplified model of Structured-light, completed the Formula Derivation of detecting parameters.(3) Combined with testing requirements of track measuring instrument, calibrated the internal parameters of vision measurement mechanism by the camera calibration method proposed by Zhang Z.Y and Matlab toolbox. At the same time, calibrated the external parameters by coinciding the light plane with calibration target, and it’s result has reached a sub-pixel level.(4) By using the difference operation of R component grayscale image and original components of the image, the characteristic line of the experimental image was accurately get. Compared the extract effect of characteristic line center by methods of morphology and improved gray gravity method. Conducted a sub-pixel extraction of the midpoint coordinates using Least-squares fitting method. And the accuracy of vision measurement system was verified by experiments.