Research Of Localization And Tracked Assembly System Based On Vision For SCARA Robots

Vision-based industrial robot is a hot spot in automatic industry in recent years,with the rapid development of computer technology and automatic control technology,industrial robots with vision occupy more and more share in automatic production line.Assembly is one of the most widespread applications about industrial robots,with the increasingly cost of human labor,it would be inefficient to do the simple assembly work via manual processes.Machine vision is becoming a key technology to improve production efficiency and ensure the quality of products.When the object is detected,the industrial robot is guided to seize the workpiece by the position information of object acquired by vision system.Aiming at these problems,the paper studied the core issues of visual orientation based with Mitsubishi SCARA robot.It mainly relates to that the robot grabs static object and matches of the target to assemble the workpieces on the conveyor belt by tracking technology.This system would achieve accurate assembly of the continuous workpiece on the conveyor belt,improve the enterprise capacity requirements and production efficiency.The main contents of the thesis are as follows:(1)This paper presents a localization and tracked assembly system based on vision for SCARA robots,and sets up the experimental platform.(2)This paper builts the camera model using the linear calibration method,and studies the conveyor belt model and SCARA robot kinematics model.This paper calculated the transformational relation among the coordinates of robot,camera and the conveyor belt by theoretical inference as well as completed their calibration.(3)According to the targeted workpieces,this paper presents the corresponding image processing algorithm.Using the Hough algorithm to detect the position of the circles and straight lines on the workpieces after pre-processing by National Instruments Vision assitant.According to the characteristic of linear motion on the conveyor belt,the paper presents a tracking process algorithm which uses the position feedback of encoder to complete shooting the workpieces without missing,and eliminate the identical workpiece with shooting repetitively.(4)The integrated control system in this paper would be developed by using LABVIEW software,the trajectory planning and control of the robot is programed based on MELFA BASIC V language.This paper applys the research results of the localization and tracked assembly system based on vision for SCARArobots to the experimental platform,then implements to grab static target and track the moving target on the conveyor belt by visual location.The experimental results show that the theoretical analysis,calibration results and control methods are correct,and the feasibility of the system positioning accuracy and the effectiveness of the assembly efficiency are effective.