Pose And Pose Measurement Of Curved Surface Based On Line Structured Light

In the field of industrial product manufacturing,the measurement and identification of the workpiece is an essential and important link.The traditional solution is usually to use manual measurement and identification of the workpiece.Under the strategic goal of "Made in China 2025",the traditional manual detection scheme can not meet the actual needs.Machine vision can make the machine more intelligent to "know" the position of things and workpiece,so that the intelligent robot to replace the artificial become the development trend of the future industrial field.In this thesis,the position and pose estimation of workpiece is studied by using linear structured light vision measurement technology.The research content includes the following five parts:(1)Build a pose measurement system.According to the requirements of the position and pose measurement system,the hardware system was built,including: Charge Coupled Device(CCD)industrial camera,linear laser,computer and displacement platform.The position of the industrial camera and the linear laser is relatively fixed.The laser projected by the laser is perpendicular to the surface of the workpiece to be measured.The displacement platform drives the workpiece to move at a uniform speed.The industrial camera collects the image with light stripe and processes it to obtain the point cloud data on the workpiece surface.(2)The calibration of the linear structured light measurement system was analyzed,the camera was calibrated with the two-dimensional checkerboard calibration method,and the calibration parameters were solved.The existing calibration methods of light plane were analyzed,and the parameters of the light plane equation were obtained by extracting the feature points on the calibration plates with different heights with light stripes.(3)Extraction of light stripe center line.Based on the analysis of the existing algorithm of light fringe extraction,an algorithm of light fringe center extraction based on the mixture of direction template and gray gravity center is proposed.The image processing technology is used to preprocess the light stripe image,and the direction template method and thinning algorithm are combined to solve the normal direction of the light stripe.Finally,the gray gravity center method weighted by Gauss is used to extract the center coordinate points of the light stripe in the normal direction.(4)The measurement method of the workpiece pose is studied.Line structured light technology is used to scan the point cloud data on the surface of the workpiece to be tested.According to the actual size of the workpiece to be tested,the workpiece model is drawn by Computer Aided Design(CAD)software and transformed into the point cloud model.Principal Component Analysis(PCA)algorithm is used to coarse-register the Point cloud on the surface of the measured workpiece and the model Point cloud to get the initial pose information of the workpiece.On the basis of the coarse-register,Iterative Closest Point is used.ICP)algorithm to complete the final workpiece pose estimation.(5)In order to verify the stability and reliability of the line structure-based light pose measurement system,this thesis selects different materials and different shapes of workpieces for experiments.The point cloud on the workpiece surface and the corresponding point cloud model were obtained by the line structured light vision measurement technology and registered to visually display the pose estimation effect.At the same time,the factors affecting the pose measurement accuracy were analyzed.