Based On Machine Vision Measurement Error Analysis And Research Of Switch Rail Size

At present,the turnout system is the key track equipment in railway operation,wh ich affects the measurement accuracy of turnout rail devices.the measurement of rail is done manually,but this method will not only cause a certain degree of error but also low efficiency,which will further affect the accuracy of the final measurement.In order to solve the above problems,one method which uses the machine vision to measure the rail is being put to use.Specifically,We use the camera in the process of measurement calibration and stereo matching technology for error analysis,developing a set of measurement system for rail parts based on machine vision model,and improving the accuracy and efficiency of measurement.First,the introduction of analyses about the source of errors in machine vision system.The source of error in machine vision system mainly due to lens distortion,projection deformation and signal conversion between the camera calibration process and stereo matching process.In this experiment,the main source of the error is the error in the measurement of rail size,and then studied the situation.Second,Aiming at the error of Harris corner extraction algorithm in camera calibration,a precise plane calibration algorithm for corner extraction is proposed.In this algorithm,the eight-neighborhood of the diagonal is analyzed locally,and the extraction of the diagonal points is optimized.After the experiment,the plane calibration algorithm reduces the error,and lays a good experimental foundation for the subsequent stereo matching.More,A stereo matching algorithm based on EC-NEC is proposed to solve the algorithm error caused by edge feature extraction in stereo matching.In this algorithm,the disparity map of edge points is obtained by feature point technique,and the non edge points are matched by region correlation algorithm under the condition of contour disparity constraint.Through experiments,the stereo matching algorithm improves the accuracy of the matching.Finally,The description of the composition and function of the whole rail measuring system.The main reason for error analysis via comparing measurement data and standard data.The experimental results show that the proposed algorithm reduces the error of the measurement system and improves the accuracy of the measurement system.