Font Size: a A A

Research And Application Of Robot Body Assembly Based On Machine Vision

Posted on:2019-07-21Degree:MasterType:Thesis
Country:ChinaCandidate:J F LiuFull Text:PDF
GTID:2438330572465384Subject:Mechanical Manufacturing and Automation
Abstract/Summary:PDF Full Text Request
In this paper,the representative assembly link of robot pedestal and RV reducer is studied,the vision is used to guide robot to grab and install,so as to complete the assembly task.The artificial assembly method which is heavy,high repeatability and high strength,has been replaced by keyboard input and click of mouse.It not only saves the manpower cost greatly,also increases the flexibility of the robot production line and make a the higher economic efficiency.This paper has mainly carried on the following research:Firstly,design the overall scheme of robot body assembly system based on vision.By using the top-down analysis method,the whole scheme is designed by the logic abstraction process of original demand-demand analysis-system design.The design and hardware selection of the sub-system are carried out according to the task demand,and complete the experiment platform construction.Secondly,identification and positioning of reducer and pedestal.The reducer and base images are processed by using median filter,the Lab color space adaptive K-means segmentation,template matching based on image pyramid and methods of subpixel edge detection of image to get image feature.Then the coordinates of bolt hole and feature hole are determined by using the least square method,and get the position and rotation of the workpiece.Thirdly,research on system calibration and grab based on Halcon.Through related operators,the camera calibration and robot hand-eye calibration are completed,and the real world coordinates are obtained using the image coordinates.Then through robot teaching,the information of grasping plane and gesture of the hand claw are obtained,and the combination of two dimensional real coordinates and these information make the research of grasping task successful.Lastly,system software development.With Visual Studio as the platform,the framework design of system software was completed by combining the method of Halcon function with the MFC application based on dialog box.Through the sub-module design,has realized thenetwork communication between PC and the robot controller,PC and workshop scheduling system,and the serial port communication and control of the camera light source,etc.,complete the development of robot assembly system of ontology finally.
Keywords/Search Tags:machine vision, robot, assembly, identification location, Halcon
PDF Full Text Request
Related items