Font Size: a A A

Research On Contact Track Detection Method Based On Structure Light Measurementtechnology

Posted on:2020-01-18Degree:MasterType:Thesis
Country:ChinaCandidate:L YangFull Text:PDF
GTID:2392330590496348Subject:Traffic and Transportation Engineering
Abstract/Summary:PDF Full Text Request
The structural light measurement technology has many advantages when it is used to detect the contact rails in subway lines.From the perspective of the industry status,the traditional manual contact measurement of the contact rail has been unable to meet the requirements of subway line maintenance due to low detection efficiency and large workload,so a portable automatic detection tool is urgently needed to provide support.From the point of technical advantages,the structural light measurement technology has great development prospect and application value because of its non-contact,strong anti-interference,and high measurement accuracy advantages which can fully meet the requirements of contact rail detection.Structural light measurement technology mainly includes two typical techniques triangulation and binocular measurement which are widely applied in the field of industrial vision measurement.The calculation amount of binocular vision is twice as much as tha t of triangulation,besides,its matching cost is high which sacrifices a certain degree of camera precision.Triangulation technology needs to get the information of optical plane structure and the requirement of its system calibration accuracy is demanding.According to the actual needs,triangulation measurement can satisfy the testing requirements through the appropriate calibration method.In this paper,basing on the actual project requirements,a triangular-based contact rail detection system is constructed to measure the geometric parameters of the subway contact rail.The main contents of the paper include the following aspects:(1)The paper analyses various factors that affect the accuracy of the visual measurement system in detail after introducing the structure light measurement model briefly.Then the in-depth study of the mathematical principles related to camera calibration,the two-step method combined with two-dimensional planar targets to achieve the calibration of the system camera.Finally,a system calibration method is designed for the measurement characteristics of the contact rail detection system and the calibration of the relationship among multiple cameras is completed.(2)To extract the structural light quickly and accurately is a difficult problem.in the structural light measurement system.Due to the influence of various environmental factors in the actual measurement process,the quality of the collected structured light images are defective.The traditional Steger method uses the convolution that bases on two-dimensional gaussian function and image to give the partial derivative of the image.Gaussian kernel parameters have a great influence on Steger algorithm,which needs to be set according to the width of structural light.Therefore,this algorithm is not suitable for images whose width of structural light changes.This paper proposes a feature extraction algorithm of structured light center in complex environment.In this algorithm,morphological algorithm is combined with filter process to extract the region with width of light band.Finally,gaussian kernel is set respectively according to the gradient of light band within the region to improve Steger algorithm.The comparison of several algorithms shows that the algorithm in this paper has strong adaptability to the environment and high extraction accuracy,especially,the extraction result is well when the algorithm is applied to the contact rail detection system.(3)The transformation from the world coordinate system to the orbit coordinate system is completed through ICP nearest point iteration.According to the contact rail characteristics in the measurement system,the feature matching between the contact rail standard value and the measured value is accomplished through slice statistics method to complete the task of feature point extraction and contact rail parameter measurement.(4)Reaching the goals of software related function design,embedded function design,and communication among softwares.According to the environmental requirements of field testing and the relationship between camera and the angle of laser position,the mechanical structure of testing equipment is designed reasonably.The test results show that the relative accuracy of the system can reach within ±0.5mm,which can meet the requirements of the measurement system.
Keywords/Search Tags:Triangulation, Contact track detection, Light strip center extraction, Iterative closest point, Slice statistical
PDF Full Text Request
Related items