Highway Guardrail Plate Parameters Detection System Based On Line Structured Light Scanning

The guardrail of highway is an indispensable safety infrastructure for highway traffic,the periodic detection of its parameters is a necessary means to ensure driving safety and reduce accident risk.The traditional manual inspection method has low detection efficiency,data is susceptible to human interference and the operating environment is dangerous,which is difficult to meet the maintenance needs of guardrail normalization.The line structured light detection technology has the advantages of high measurement accuracy,fast speed and noncontact.Therefore,this thesis designs a highway guardrail plate parameters detection system based on line structured light scanning,in order to realize the automatic inspection and digital filing of the collision guardrail parameters.This thesis takes the parameters detection of the W-beam guardrail plate as the research object,and adopts the intelligent inspection robot carrier platform independently developed by the team.The key technologies such as the composition scheme of the inspection robot system,the image processing of the strong reflection surface light stripe,the geometric parameters detection algorithm of the guardrail plate,the calibration of the sensor and the robot system,and the design of the digital archiving and management platform are mainly studied.The efficient and automatic detection of the multi-dimensional parameters of the guardrail plate is realized,which has important theoretical significance and practical value for the maintenance and performance evaluation of the guardrail facilities.In the process of the research:Firstly,the overall scheme of structured light scanning detection based on autonomous inspection robot platform is designed for system detection requirements,and the structure of structured light vision sensor is optimized according to the optical path principle of direct laser triangulation.By analyzing the camera projection imaging model,the robot detection system model is established,and the conversion relationship between the pixel coordinate system and the robot coordinate system is determined.Furthermore,the corresponding calibration model is established by using the spatial attitude angle of the robot to compensate the error introduced by the robot attitude deflection.Secondly,by analyzing the strong reflection characteristics of the guardrail surface and the characteristics of the light stripe image,an adaptive exposure algorithm is designed to suppress the image highlight,and the image preprocessing and the light stripe specular reflection removal algorithm are further realized to filter out the background interference and obtain the complete light stripe target.Then,aiming at the problem of feature point extraction of light stripe,the condition of determining feature points by traditional slope threshold method is improved to accurately identify the feature points of two kinds of guardrail plate light stripe.On this basis,combined with the center coordinates of light stripe,the detection algorithm design of guardrail plate shape,deformation,inclination and center height is completed.In addition,the sensor and robot system are calibrated by using the checkerboard plane target,the parameters in the mathematical model of the system are determined,and the data output of the two-dimensional pixel coordinates in the robot coordinate system is obtained.In order to meet the actual needs of online detection and measurement data archiving,the development of visual inspection software and digital archiving management platform is completed.Finally,a prototype of the system is built,and the prototype is used to carry out the relevant experiments of the detection system,which verifies the feasibility,stability and accuracy of the structured light scanning parameters detection system.The experimental results show that the system can efficiently and accurately detect the shape,deformation,inclination and center height parameters of guardrail plate.