Research On Robot Intelligent Handling System Based On Machine Vision

With the rapid development of the intelligent manufacturing industry,more and more industrial robots are applied to the process of industrial production.The development of industrial robot intelligent transport system instead of traditional teaching control system is the trend of automatic production equipment,usingthe machine vision technology to extract the image features and the automatic handling of the box is realized.For the application of vision to the robot control system,the main research work is as follows:Firstly,designing a control system of box handling robot based on vision,and analysising the kinematics of the MOTOMAN-UP6 six axis robot,using the Descartes coordinate equation to solve the positive and inverse solutions of the kinematics of the six axis robot;Designing an electric powered mechanical hands for box handling and using UG software for three-dimensional structure modeling,and then making machine tool prototype with referenceto structural parameters;designing and making a mechanical hand controller according to the function of the mechanical hand,and optimizing the controller circuit so that the control and communication of the mechanical hand controller can be more safe and reliable.Secondly,for the software development process of the handling robot control system,according to the underlying requirements of development of Open CV image processing software,MER-Series industrial camera software,MOTOCOM32 robot control software to configure in Visual Studio programming software;The visual interface of robot control and the image processing interface are developed,and the real-time control of the MOTOCOM32 robot control library function is carried out to realize the robot movement and the real-time control of the mechanical hand.Thirdly,selecting the MER-130-30 UM industrial camera and calibrating the parameters of the industrial camera according to the calibration algorithm of Zhang Zhengyou camera;Using the OTSU algorithm to improve the adaptive threshold of the Canny edge detection algorithm and solve the influence of the external light intensity on the edge processing quality;Using the Hough transform algorithm to obtain the attitude angle parameters of the box in the image,and using the Hu moment algorithm to obtain the coordinate parameters of the box.Finally,building an intelligent handling system for control verification,and the actual position and position parameters of the box are obtained by using the image parameters processed by the image and the coordinate transformation of the camera calibration parameters,and then the control system automatically plans the moving path to complete the robot's automatic handling of the box;Analysing the errors between the position and pose parameters of the coordinate conversion box and the ones of the actual measurement to make the intelligent handling system more safe and reliable.