Research On Measuring Method Of Rail Head Wear Based On Line Structure Light

As a key part of railroad tracks,the railroad tracks bear all the forces directly between the vehicle and the rails.Its working conditions are very poor.Severely worn rail heads can affect wheel-rail contact conditions,causing periodic impact on vehicle body.At the same time,it will have a negative impact on railway equipment and transportation..The current inspection and measurement of rail wear in China is still dominated by manual inspection methods.This kind of maintenance is Inefficient and the inspection results and maintenance standards are greatly influenced by the subjective factors of maintenance personnel.Therefore,it is of great significance to develop an efficient and accurate railhead wear measurement technique for maintaining the state of railway track and ensuring the safety of traffic.Under the support of the National Natural Science Foundation of China “High-speed machine wheel rail dynamics computer vision experimental research”,this paper proposes a rail-based track head wear measurement method based on image processing technology and line structure light measurement technology.The main work are as follows:Firstly,by comparing and analyzing various methods for measuring the wear of rail heads at home and abroad,the basic scheme of rail head wear measurement based on line structure light was selected.And this paper designs the calculation method of the rail profile registration feature points and the wear calculation feature points based on the geometric characteristics of the rail profile.Secondly,according to the measurement requirements of rail wear,this paper has completed the selection and construction of the rail outline image hardware and software processing platform.At the same time,a convenient calibration method for structured light system is designed based on Zhang's method,which realizes the conversion of image coordinates and camera space coordinates.It is proved by experiments that this method has high accuracy and can meet the requirements of measurement.Then,in view of the problem that the structured light bar image in railway applications is easily disturbed by the environment,this paper makes an modified retinex algorithm based on lab space.In low illumination and haze environment,the structured light stripe image can be significantly enhanced by this algorithm in color and texture.It is beneficial to the following process of image processing and rail contour reconstruction.Thirdly,this paper uses the image processing technology to realize image segmentation,filtering and denoising in the channel R.Then,the improved Hessian matrix light strip center extraction algorithm is used to extract single-pixel rail profile curve,which has high operational efficiency.Finally,the calibration and measurement results are used to calculate and evaluate the rail profile curve and wear.By comparing and analyzing the measurement parameters of different types of standard rails and wear rails with manual measurement results,the stability and accuracy of the method have been verified,which can meet the application requirements of non-contact rail head wear measurement.