Research On The Manipulator Control System Based On Stereo Vision

Stereo vision technology can obtain the depth information of the target directly and It is widely used.By combining stereo vision technology with industrial manipulator control system,the mechanical arm's intelligence and automation have been improved steadily.The robot arm control system based on stereo vision has been paid much attention by the academic and engineering circles,and this phenomenon has important theoretical and practical significance for its research.This paper is based on the type of IRB120 six degrees of freedom industrial manipulator which made by ABB industrial.The recognition and positioning of small blocks are accomplished by stereo vision technology and the grasping action is completed by mechanical arm control technology.The main contents are as follows:First,the camera calibration and image filtering algorithm are studied.Firstly,the imaging principle of stereo vision is introduced.Then,Zhang's calibration method is used to solve the internal and external parameter matrix of the camera.Finally,considering that the image will be affected by noise in the process of image acquisition,the image filtering algorithm is studied.By comparing the experimental results of different filtering algorithms,adaptive median filter is selected to filter the target image.Second,the target recognition and localization method of small building blocks is studied.The quality of feature extraction determines the subsequent recognition effect.Therefore,feature extraction algorithm is studied.Firstly,an optimized corner detection algorithm is proposed by introducing pre-screening and adaptive threshold to optimize the shortcomings of traditional Harris algorithm.Then the SIFT algorithm is introduced.According to the advantages and disadvantages of the two algorithms,this paper combines the optimized Harris and SIFT,and takes the feature extraction algorithm as the feature of template matching to complete the target recognition.Finally,the edge based center solution algorithm is used to locate the identified object center.Thirdly,the visual control method of the manipulator is studied.Firstly,a D-H parameter model was established according to ABB's IRB120 mechanical arm structure characteristics,and the forward and inverse kinematics were analyzed based on this model.Then the hand eye calibration method of camera and manipulator is done.Next,the interpolation algorithm of the manipulator is used to control the tracjectory of themanipulator to grasp the target object.Finally,Robotstudio software is used to simulate the grasping action of the manipulator.Fourthly,the robot arm experiment platform based on stereo vision is built.The experiment of manipulator positioning and grasping is carried out by using the method mentioned above.The experimental results show that the method adopted in this paper can realize target recognition and positioning,and control the manipulator to complete the grasping action.