Research On The Technology Of Robot Positioning And Grasping Based On Binocular Vision

Industrial robots have a long history of development while the combination of industrial robots and vision is an important manifestation of intelligence.With the development of machine vision,monocular vision has been developed from the first to nowadays binocular,multi-vision.Because binocular stereo vision can obtain the three-dimensional spatial information of the target object by parallax and can adapt to different environments,it has the unparalleled advantages of monocular vision.In this paper,binocular stereo vision system is set up by using Yaskawa MA1440 industrial robot to study the key technology of binocular positioning based on binocular vision.First of all,this article briefly introduces the background and significance of the topic selection,elaborates the research status of binocular vision robot at home and abroad.The working principle of the camera,the linear and nonlinear imaging models are studied.Based on the principle of binocular vision,a binocular vision manipulator positioning and grasping system is designed and built,which mainly includes the selection of industrial manipulator and fixture,design and installation,the selection of industrial cameras,lens and image capture card,the binocular Distance,angle calculation,and three-dimensional modeling of the binocular vision system.Secondly,aiming at the system calibration,the principle of monocular camera calibration is studied.Different calibration methods are compared.Halcon calibration method is used to calibrate the binocular camera.The structural parsameters of the binocular vision system are obtained through calibration experiment.Experiments show that the accuracy of the results obtained through the improved calibration method is higher.Researching robot hand-eye configuration model,a method of rapid calibration of hand-eye is proposed and implemented,and finally hand-eye calibration parameters are obtained.Next,the image enhancement algorithm of gray-scale transformation and image smoothing filtering is studied for the image acquired after calibration,an improved hybrid image enhancement algorithm is proposed,and the effectiveness and robustness of the algorithm are verified through experiments.In addition,the principle and method of stereo matching are researched and elaborated.The stereo matching method based on region is selected and improved.The experimental results show that the improved stereo matching algorithm has good matching accuracy and high efficiency.Then,the principle of 3D reconstruction is expounded in detail and the accuracy and stability of the system reconstruction are verified through reconstruction experiments.According to the grasping of scattered workpieces,an image segmentation algorithm based on region growth is realized,and the highest target workpieces scattered randomly are extracted.The model-based three-dimensional pose estimation method is studied,and the three-dimensional pose estimation of the target workpiece is realized.Finally,a binocular vision robotic positioning and grasping overall design scheme is proposed,and a set of binocular vision positioning system integration software is designed and developed.Based on this,experimental results show that this binocular vision manipulator positioning and grasping system has recognition Fast positioning,high precision crawling and practical advantages.