Research On Key Technologies Of Three-dimensional Measurement Based On Line Structure Light

With the development of the manufacturing industry,the traditional contact measurement method can not meet the requirements of speed and precision of industrial production.Visual measurement has received extensive attention as a non-contact,real-time 3D measurement method.The three-dimensional measurement technology of line structure light based on the principle of triangulation is one of the research hotspots in this field because of its high precision and large scale.How to accurately calibrate the line structure light vision measurement system and extract the center of the light stripe in complex environment is the focus of this technology research,which is of great significance for improving the flexibility of the system and the accuracy of measurement.This paper takes the line structure light three-dimensional measurement system as the subject of the research.The key problems such as calibration of vision sensors and center extraction of light stripe in the system are studied.Finally,a visual measurement platform based on line structure light was built.The measurement accuracy of the system was evaluated by height measurement of the standard gauge and 3D reconstruction of the surface.The main research contents of this paper can be listed as follows:(1)The basic principle and mathematical model of 3D vision measurement based on line structure light are studied,and a 3D vision measurement platform based on line structure light is built.(2)Aiming at the calibration problem of visual sensors,the camera calibration technique and the structured light-plane calibration method based on the principle of cross-ratio are studied.The structured light vision sensor was calibrated using a modified two-dimensional checkerboard target and the accuracy of the calibration was evaluated.(3)Aiming at the problem of center extraction of complex structured light stripe on metal surfaces,a suitable light stripe center extraction algorithm is proposed.Firstly,the light stripe is segmented from the complex background environment by combining the gray and gray gradient information of the light stripe in the light stripe image;then the binary image area constraint is introduced to obtain the ROI light stripe region;finally,the gray-gravity method is used to obtain the center of the light stripe,the effectiveness of the algorithm is verified by experiments.(4)According to the constructed visual measurement platform,the heightmeasurement of the standard gauge and the 3D reconstruction of the surface were completed.The accuracy of the system was evaluated,the cause of the error was analyzed,and the surface 3D reconstruction of some objects was carried out.