Research On Optical Path Calibration Method Of Track Settlement Monitoring System Based On Machine Vision

In the process of railway operation,when the track settlement or deformation occurs,it may lead to serious accidents.Therefore,it is necessary to grasp the track settlement and deformation in time to ensure the safe operation of the railway.Compared with traditional methods,image-based track settlement monitoring technology can solve these problems more effectively.Image processing technology is used to realize large-scale,automatic,non-contact,real-time and high-precision settlement monitoring of monitoring sections.In this paper,it is difficult to manually adjust the laser spot to the target center of the adjacent system due to the long distance between the monitoring equipment during the operation of the image-based settlement monitoring system.In the monitoring process,when the track settlement is too large,it may cause the spot to deviate from the field of view of the camera,resulting in settlement monitoring failure or data error,so it is necessary to calibrate the optical path.In view of the above problems,this paper proposes an optical path calibration method for track settlement monitoring system based on machine vision.Firstly,the error between the current position and the expected position of the laser spot is obtained by image processing technology.Then,the stepper motor is controlled by the calibration system to adjust the laser position to achieve the laser spot to reach the specified position.Therefore,this paper focuses on the two aspects of laser spot center positioning and stepper motor control algorithm.The main work is as follows :1、According to the requirements,the overall framework of the improved spot center positioning is designed,and the appropriate hardware and development environment are selected according to the performance requirements.In order to improve the accuracy of the spot center positioning,the camera is first calibrated to obtain the internal and external parameters of the camera,reduce the image distortion generated by the camera acquisition process,and improve the accuracy of the subsequent spot center positioning detection results.Then,the traditional Canny edge detection operator is improved.The 3 * 3 finite difference template is used to replace the traditional 2 * 2 finite difference template,and the gradients in45 ° and 135 ° directions are increased.The OTSU algorithm is used to find the high threshold in the Canny operator,and the 1 / 3 of the high threshold is used as the low threshold of the Canny operator.The morphological processing is added to reduce the internal noise of the image.By comparison,the improved Canny operator in this paper has higher detection accuracy.Finally,the traditional gray center of gravity method is improved.The local weighted gray center of gravity method is used to extract the region of interest according to the edge pixel coordinates obtained by the improved Canny operator.Only the weighted gray center of gravity is calculated for the region of interest to reduce the amount of computation and interference factors.Through experimental verification,the improved algorithm proposed in this paper has higher accuracy and can accurately locate the spot center.2、The error of the current position and the expected location of the outlet of the outlet of the image,a calibration mechanism is required to adjust the laser position.This paper designs a calibration scheme,step motor is used as adjusting device,the choice of stepping motor.In this article,two phase hybrid stepping motor is chosen to study,and a mathematic model is set up,and then discusses the traditional PID tax algorithm and fuzzy tax theory.In this article,an adaptive fuzzy PID controller is proposed as the control algorithm,and two control algorithms are developed.3、Through MATLAB software,the traditional PID control algorithm and the adaptive fuzzy control algorithm designed in this paper are simulated and verified.The simulation results show that the adaptive fuzzy control algorithm used in this paper has stronger adaptive ability,higher tracking accuracy than the traditional PID control,and the response time is also smaller than the traditional PID.The performance of the adaptive fuzzy PID control algorithm used in this paper is verified.4、Firstly,the hardware implementation of the spot center positioning system is carried out.Through the traditional gray center of gravity method and the improved algorithm in this paper,multiple sets of data are collected respectively.After verification,the error of the improved algorithm in this paper is much smaller than that of the traditional center of gravity method in the actual experiment process,which verifies the effectiveness and superiority of the spot center positioning algorithm in this paper.Then the optical path calibration is simulated and verified,and the optical path calibration in the initialization stage and the optical path calibration in the monitoring stage are simulated respectively.It is concluded that the adaptive fuzzy PID used in this paper has higher tracking accuracy and smaller error.