Research On Intelligent 3D Vision Measurement Method Of Robot Based On Line-structured Light

In recent years,3D visual measurement technology has been widely used in the field of modern industrial manufacturing and product quality inspection.With the improvement of measurement accuracy and efficiency,3D visual measurement technology has become the main means of non-contact measurement method,and will also become the mainstream trend in the field of industrial inspection in the future.In3 D measurement of large,complex morphology objects,traditional the combination of the line-structured light sensor and the displacement of the platform as a result of the limitation of working space and pose unable to realize the effective measurement on the surface of the object,combines industrial robots and line-structured light sensor to eliminate this limitation,realize arbitrary position of objects on the surface of the 3D scanning and detection.However,in the 3D scanning process of the object,due to the complex surface topography of the measured object and the limits of the sensor range,it is difficult to plan the 3D scanning path of the robot,making the operation of the measuring equipment complicated.Therefore,it is very important to research a method that can intelligently plan scanning path for robot 3D measurement system based on line-structured light vision.In view of the above problems,this paper improves the traditional 3D measurement system combining robot and line-structured light,and proposes an intelligent measurement method based on depth camera perception to solve the problem of 3D scanning trajectory planning of complex shapes objects,and further improve the measuring efficiency of equipment.In-depth theoretical and experimental research is conducted on measurement system model,point cloud data processing,and scanning trajectory planning.The main research contents are as follows:Firstly,the mathematical model of the line-structured light measurement system of robot was analyzed.The forward kinematics model of robot was established through the modified D-H method,and the modified D-H parameters of each bar of the 6-dof industrial robot were calculated,and the mathematical model of the line-structured light sensor was analyzed.Eye-in-Hand calibration method was researched.The spherical coordinate of standard ball was taken as a fixed point in space,and the robot was controlled to carry out four times of translation and six times of posture movement measurement,and the coordinate transformation matrix from the robot terminal coordinate system to the sensor coordinate system was calculated.PCL was used to analyze and research the processing of scanning point cloud and3 D reconstruction algorithm.The sparse abnormal values were eliminated by statistical filter,and the data were sampled by voxel to complete the pre-processing of point cloud data.The point cloud data index of KD-tree and the calculation method of point cloudnormal vector are analyzed.The greedy projection triangular surface reconstruction algorithm is used to realize 3D reconstruction.The scanning trajectory planning method of Kinect perception was studied,and the measurement principle of depth camera was analyzed.The Eye-to-Hand calibration algorithm is researched and analyzed theoretically to calibrate the pose relationship between the base coordinate system and the depth camera coordinate system.A robot scanning path planning method based on the point cloud data of the depth camera is proposed.Path points are calculated by point cloud direct filtering,point cloud plane segmentation and point cloud slice.Design and implement the measurement system proposed in this paper,complete the construction of the system platform,write the data acquisition software of line-structured light sensor,and the data acquisition program of point cloud,and realize the synchronous acquisition and processing of robot joint Angle data and sensor contour data,and display the point cloud.Through PCL,the algorithm of point cloud data processing and 3D reconstruction is implemented and compared with commercial software.The validity of 3D scan path planning method based on depth camera perception is verified by experiments.By comparing the point cloud data obtained from the experiment with the CAD model,the accuracy analysis shows that the measurement standard deviation of the scanning data is better than 0.46 mm,which verifies the effectiveness of the measurement system proposed in this paper.