Delta Robot Wine Bottle Pinball Sorting System Based On Machine Vision

In the actual industrial production,the workpiece sorting on the assembly line is an indispensable component unit in the production process.The traditional manual sorting method has low efficiency and high operation intensity,which increases the production cost in a disguised way.Compared with this operation mode,the machine vision and industrial robot intelligent sorting method can analyze and control the workpiece in real time,and has the advantages of fast,stable,efficient and high accuracy.According to the demand of the current market,the automatic intelligent sorting system of workpiece based on machine vision has a very wide prospect in the field of industrial application,and has good practical value and deeper significance.In this paper,based on machine vision and Delta robot,the improved QPSO-BP algorithm is used to realize the identification and positioning of the wine bottle on the conveyor belt,and the precise coordinates of the workpiece are obtained.Finally,the robot picks up the marbles from the marble chute and places them on the wine bottle mouth.The main research contents are as follows:1.Design the 3D model of the whole machine,the control method of the vision ystem and the robot control system are designed.The equipment needed for he vision system was selected and purchased,and Lab VIEW graphical rogramming software was used as the upper computer software of the system. CP/IP communication protocol was used to establish the connection between he robot and the vision system,so as to transmit coordinate data in real time.2.The forward kinematics and inverse kinematics of the robot are calculated and wo sets of solutions are obtained.At the same time,the workspace boundary f the robot is calculated and simulated in the software to ensure not conflict ith marbles chute and smooth operation of the robot in the space.3.Carry out research on visual system,including several coordinate system ransformation,camera imaging principle.A hand-eye calibration method is sed to calibrate the corresponding relationship between the camera and the nd of the Delta robot,and the calibration method of the motion direction vector etween the robot and the conveyor belt is also include.Finally,the collected mage of wine bottle mouth is preprocessed by image enhancement,filtering nd denoising,and the target image can be directly used for positioning.4.The image was roughly positioned by template matching algorithm based on entroid and circular edge fitting algorithm based on edge,and then accurately ositioned by improved QPSO-BP algorithm,the accurate coordinates of the orkpiece were obtained,and then sent to the robot through coordinate system onversion.After a period of trial operation,the positioning method is stable and reliable.5.The machine is assembled into a whole according to the mechanical design drawings and electrical principle construction drawings,and then the hardware and software part of the debugging experiment.Hardware,mainly related to trigger the bottle sensor,marble chute and shaker reasonable distribution.The software part includes writing visual program and encoder tracking program.In the end,on-line debugging,the experiment to get the whole process of marble sorting.And continuous trial run experiments to ensure that it can meet the requirements of industrial application.