| With the rapid development of railway in China,the loss of rail infrastructure is serious.The condition of rail fastener is directly related to traffic safety.Therefore,it is very important to study an accurate,efficient and safe fastener detection method.In this paper,a3 D geometric detection method of rail fasteners based on multi-line structured light 3D scanning technology is proposed.A fastener detection system is designed for fastener defects and verified by experiments.The main work contents are as follows:(1)Establishment of a structured light 3D scanning modelVisual measurement model is an effective method to realize 3D scanning.It is necessary to analyze and process the feature information of 2D image to realize the measurement of 3D surface information.Traditional binocular vision can not meet the requirements of industrial detection because of the sensitivity of environmental light.Structured light vision uses laser projectors to project light to intersect the surface of objects,and obtains three-dimensional coordinate values of objects through two-dimensional light strip information,which has the advantages of large field of view,easy extraction and real-time accuracy.The monocular multiline structured light 3D scanning model proposed in this paper is composed of multi-line laser emitter,industrial camera and real-time dynamic processing software,which can obtain the3 D point cloud information of fastener surface in real time,with high extraction speed and accuracy,and can meet the requirements of fastener detection.(2)Fringe center extraction of structured light and field calibration of multi-line structured lightAccording to the commonly used algorithms for extracting the center of structured light fringe,the differences in extraction effect and rate are compared,and finally the gray barycenter method is selected as the algorithm for extracting the light strip in this system experiment.In the field calibration of multi-line structured light,the main process is to determine the twodimensional target plane equation,extract the center point of the optical fringe and classify it,then construct the target plane equation through the center of light and the ray equation of the optical strip point in parallel,obtain the three-dimensional coordinate points of the optical fringe by combining the principle of line and surface intersection,shoot multiple images to obtain the three-dimensional coordinate points of the optical fringe,and fit the optical plane equations by the least square method.(3)Fastener elevation information coding and defect status detectionIn this paper,a defect judgment method is proposed to accurately judge the defect type by comparing the actual measurement results with the standard coding.After obtaining the three-dimensional point cloud information of fasteners,the distance between the point and plane plane is solved by using the point-plane distance formula,and then the three-dimensional elevation data of fasteners is obtained.These data were divided into a grid of 32/16/8×5×4according to different elevation values,and specific 32/16/8 elevation values were selected to encode fastener information.The type of defect can be determined and the state of fastener can be detected by comparing the three-dimensional elevation coding with the static standard coding.Through the above work,a 3D geometric detection method for rail fasteners based on multi-line structured light has been successfully implemented,which can meet practical inspection requirements and has achieved good experimental results. |
