3D Measurement Of Complex Workpieces Based On Depth Camera

In the field of industrial spraying,with the continuous development of target measurement technology of machine vision,rapid and accurate three-dimensional measurement of workpiece can do a good job for the trajectory planning of spraying robot,reduce the uncontrollable impact of manual operation on spraying quality,and realize spraying automation in enterprises.However,there are many problems in spraying application scene,such as complicated structure of workpiece,dust and vibration interference,complex background,etc.Traditional machine vision measurement methods are difficult to meet the requriments of actual industrial spraying automation application.Therefore,how to find an efficient and high-precision threedimensional measurement method for workpieces on the assembly line of spraying enterprises has extremely important research significance.In order to obtain the three-dimensional dimension of the workpiece and realize spraying automation,a three-dimensional measurement scheme of complex workpiece based on depth camera is proposed in this paper.The three-dimensional reconstruction of the spraying workpiece is realized,and the coordinates and size information of the workpiece are obtained through the coordinate mapping,which can guide the spraying robot to do the spraying work better.The main research contents are as follows:1)The characteristics of spraying pipeline and workpiece to be measured were analyzed.Kinect depth camera with strong anti-interference ability,high measurement accuracy and high sensitivity is used as imaging device,and the experimental platform for three-dimensional reconstruction measurement and imaging system was designed and built.Bsed on the calibration theory,the Kinect camera was calibrated to obtain the corresponding relationship between the camera and the real world coordinate system,providing a unified coordinate reference system for the acquisition of point cloud and the calculation of three-dimensional coordinates.2)The imaging principle of Kinect depth image was studied and the depth image of workpiece was collected.The background removal method was proposed based on complex factory environment.After denoising,the depth image was transformed into point cloud image through object-to-point transformation,and the image was simplified to improve the efficiency of point cloud registration and fusion in the later stage.3)Registration and fusion of workpiece point clouds.ORB(Oriented FAST and Rotated BRIEF)algorithm is used to extract feature points of workpiece image,and KNN(K-Nearest Neighbor)and RANSAC(Random Sample Consensus)algorithm were combined to remove mismatch.After obtaining two-dimensional feature matching points,the mapping relationship of three-dimensional points was obtained through coordinate mapping,and then the initial transformation matrix is obtained to provide a more reasonable initial iteration value for ICP(Iterative Closest Point)registration algorithm.The registration efficiency of point cloud is effectively improved.Then,the truncated symbol distance field algorithm is used to fuse the registration point cloud data,and the three-dimensional reconstruction model of the workpiece is obtained.4)The software platform of workpiece three-dimensional measurement is developed by using C++ programming language,and the three-dimensional measurement experiments of various workpieces are carried out under the actual environment of the factory.The results show that the three-dimensional measurement system built in this paper can effectively restore the contour of the three-dimensional workpiece,and the measurement accuracy can reach ±6mm,The obtained threedimensional data can effectively guide the spraying robot to plan the related spraying trajectory.The three-dimensional measurement system of spraying workpiece will help to improve the automation and intelligence of small and medium-sized spraying enterprises.