Rotary Scanning Three-dimensional Measurement System Based On Line Structured Light

Three-dimensional measurement technique is one of the important methods to obtain the surface shape characteristics of the measured object.In recent years,with the development of computer technology,the 3D measurement scheme based on computer vision has been widely used in many fields,such as cultural relic protection,industrial measurement,medical diagnosis and so on,for its non-contact,high precision and high efficiency.Based on the principle of laser triangulation,this paper establishes a three-dimensional measuring system of rotating scanning line structured light.Through the scanning of line structured light,the 3d coordinate information of the surface contour of the measured object is collected,and the 3d digital model of the measured object is established.Finally achieve the purpose of three-dimensional measurementFirst,by analyzing various 3D measurement solutions,This paper proposes a rotating scanning method to obtain point cloud data on the surface of the measured object.Aiming at the problem of blind spot detection in monocular structured light measurement system,the binocular structured light vision sensor is constructed by adding cameras.In addition,by analyzing the mathematical model of the rotating-scan 3d measurement system,the paper establishes the conversion relationship between camera plane and linear structured light plane,and determines the parameters of the vision detection system according to the system calibration experiment.Secondly,this paper use MATLAB to simulate a variety of different Laser stripe center extraction algorithms,and comparing the advantages and disadvantages of various algorithms.Finally,a sub-pixel Laser stripe center extraction algorithms based on the direction field of the structured light stripe and the bilinear interpolation method is proposed.This method not only guarantees the precision of the Laser stripe Central extraction,but also meets the real-time requirement of the system.Third,because point cloud data has the characteristics of large data volume and uneven distribution.Before building 3D model of the object based on point cloud data,this paper first uses filtering algorithm and point cloud simplification algorithm to preprocess the point cloud data.In addition,the paper accelerates the algorithm by establishing the topology of the point cloud data.On the basis of pre-processing of point cloud data,the paper improves the ICP algorithm to fuse the point cloud data collected by the two cameras.Finally,the surface model of the object is established by surface reconstruction algorithm.Finally,the paper builds a hardware platform for the 3D inspection system and designs visualization software.Then by establishing the 3D digital model of the object,the 3d measurement of the object is completed,and the reliability of the 3d detection system proposed in this paper is verified.