Design Of Industrial Robot Positioning System Based On Improved Algorithm Of Point Cloud Registration

With the upgrading of our country's industry,the assembly line operation of factories has higher requirements for automation and intelligence.For industrial robots,machine vision to guide operations is becoming more and more popular,and enterprises' demand for threedimensional vision guided robots is increasing day by day.Three-dimensional vision can solve problems that cannot be solved by traditional two-dimensional vision,and has the potential to replace traditional two-dimensional vision.The most critical problem in the 3D vision system is to determine the position of the target object.In order to solve this problem,this paper designs an industrial robot positioning system based on an improved algorithm for point cloud registration.The main work done is as follows:(1)An industrial robot positioning system based on improved point cloud registration algorithm is built to realize the function from acquiring the point cloud of the target scene to the robot positioning the target object.This article chooses to use the KUKA robot KR-5-R1400 as the action execution part of the industrial robot.The vision system uses structured light 3D scanning technology that can balance accuracy and speed.This article chooses to use Texas Instruments DLP4500 projector and gray point camera FL3-U3-13Y3 M to build an environment for obtaining point clouds.(2)Use Gray code combined with phase shift technology to build a structured light threedimensional scanning system.The inter-zone phase expansion method is used to effectively eliminate the edge jump phenomenon caused by the misalignment of the boundary between the Gray code and the phase shift projection.Finally,the singular value decomposition method is used to solve the hand-eye calibration problem from the point cloud coordinate system to the robot coordinate system.Perform camera calibration and projector calibration to obtain the internal parameters,external parameters and distortion parameters of the camera and projector.The camera and projector are paired to establish a structured light 3D scanning system.This article uses Gray code combined with phase shift technology to perform structured light projection,which shows good accuracy and speed.(3)The improved ICP registration algorithm is used to obtain the precise pose of the target object.On the basis of the classic ICP algorithm,uniform downsampling is used to streamline the input source point cloud data,the Kdtree point cloud data structure is used to speed up the search speed in the process of finding the closest point of the target point,and the linear least square method is used to optimize the transformation matrix solution process.The target object template point cloud is registered with the scene point cloud to obtain the pose transformation of the target object in the current scene,and then the pose is transformed to the robot coordinate system through hand-eye calibration to obtain the position of the target object in the robot coordinate system.(4)Complete the construction of industrial robot positioning system and design experiments.By using standard point clouds and experimental artifacts to perform registration experiments,and analyzing experimental data,it can be concluded that the improved ICP algorithm greatly shortens the registration time without affecting the accuracy.Use industrial robots to perform positioning experiments on the moved target object.The results of the positioning experiment show that the industrial robot positioning system based on the improved point cloud registration algorithm can accurately locate the target object.