The Research Of Technology And Experiment On Robot Picking Scattered Workpiece Based On Line Structure Light

The picking technology of robot is the premise of automated loading,unloading and assembly of workpieces by using the robot,and the pose measurement of the scattered workpiece is the key technology for the robot to automatically pick up.The machine vision system can automatically collect the scene information of scattered workpiece,and complete the pose measurement of the workpiece,which provide the possibility for the robot to pick up the scattered workpiece automatically.Therefore,the research on the technology of picking scattered workpieces by robot based on machine vision has an great significance to solve the problems of loading,unloading,assembling of scattered workpieces,and improving the efficiency of industrial production.In this paper,the measurement technology of line structure light vision is applied on the pose measurement of scattered workpieces.A set of the picking system for scattered workpieces is designed and developed,and experimental research are carried out.The main research contents and results of this paper are as follows:(1)The definition and pose measurement method of the picking system for scattered workpieces are expounded.By studying and comparing several kinds of three-dimensional visual inspection methods at present and their advantages and disadvantages,the method of line structure light vision is selected according to the measurement requirements of the system.The model of three-dimensional measurement of line structure light vision is analyzed.The planar template method is used to calibrate the picking system for scattered workpieces.The calibration principle and algorithm of different parameters are introduced.(2)A pose measurement algorithm for scattered workpieces is designed.Firstly,the collected light strip image is processed,an improved algorithm of strip center extraction is proposed based on the analysis of different algorithms of strip center extraction.Secondly,the image segmentation algorithm based on region growing method is used to obtain the workpiece to be picked up.Finally,the surface point cloud of the workpiece is reconstructed,the point cloud of the workpiece and CAD model are matched by the matching method from HALCON to calculate the spatial pose of the target workpiece.(3)To develop a picking system for scattered workpiece based on line structure light.On the basis of the overall scheme design of the system,the hardware of the system is analyzed and selected,and the hardware platform is built;On the basis of the function analysis of system software,the design of the software module and human-computer interaction interface is completed through MFC and HALCON.(4)To test and verify the developed picking system for scattered workpiece.The parameter calibration experiment of the line structure light vision system and the robot hand-eye relationship is first carried out,and the relevant parameters of the the developed picking system are obtained.Then,tee and elbow were selected as experimental object to carry out experiments to verify the feasibility of the developed picking system for scattered workpiece.