Research On Real Time Positioning And Grasping Of Static/Dynamic Targets With Single Manipulator And Binocular

In the "2025 strategic plan of Chinese manufacturing",the robot as one of the strategic plan ten to give priority to the development of industry,the potential competitiveness of its theoretical depth and breadth of the market and are not to be underestimated.As the most important unit of the robot,the robot arm is widely used in the fields of industrial production,aerospace,military and daily life.Different from the early teaching mode,more and more sensors are used to assist the work of the manipulator,which increases the flexibility and autonomy of the manipulator.Visual as the most important source of information is widely used in cooperation with manipulator,constitute the visual servo system,that is,hand-eye system.The hand-eye system is used for grabbing the target object through visual system parameters calibration,target recognition detection and positioning,manipulator motion model establishment and motion control algorithm and so on.Each link has its own difficulties,the realization is complex,and the research significance is profound.Firstly,clearly the coordinate system and its conversion relation of the visual system,using Matlab Calibration Toolbox calibrating The internal parameters and external parameters of Bumblebee2 binocular camera,and analyze the calibration results.The detection and recognition of static objects(based on contour detection and color recognition)are carried out after image prepossessing(Image gray,denoising,threshold segmentation,mathematical morphology)of static target.The object of the left and right view is matched by the dot matching matrix and the sequential consistency constraint criterion.Based on the principle of binocular positioning,the objective object 3D position of the space based on the left camera coordinate system are calculated(this paper does not consider the attitude constraint),and the distance measurement experiment is carried out to verify the accuracy of positioning.Secondly,the detection and tracking methods of moving objects are studied.The advantages and disadvantages of the inter-frame difference method,the background subtraction method and the optical flow method applied to the motion target detection are discussed and analyzed.This paper introduces an improved target tracking method based on mean shift theory.Combining with the target detection and tracking method,the Camshift improved algorithm based on background subtraction ROI region automatically generated is proposed,and the moving object detection is analyzed by VC6.0 video display module in the case of occlusion.The binocular positioning formula is used to calculate the target centric coordinates of each frame,and the three-dimensional coordinate trajectory is recorded.Thirdly,based on the six-degree-of-freedom rotating industrial robot configuration,the kinematic model is analyzed by D-H method.Combined with the manipulator geometry analysis,the improved Pieper algorithm is used to solve the inverse kinematics.This paper analyzes the trajectory planning method of the 6-DOF manipulator for the straight and quadratic quadratic spline curve,analyzes the reachable space based on the special gripping attitude of the manipulator,constructs the trajectory planning experiment platform with VC6.0,and carries on the experiment to the linear trajectory planning.The reachability of the special grab attitude of REbot-V-6R 6-DOF manipulator is simulated by Matlab toolbox.Based on the eye-to-hand configuration model,the traditional hand-eye calibration method is proposed.Combine the camera coordinate system with the robot-based coordinate system,the transfer matrix is solved based on the fixed-point motion of the manipulator.Finally,the design of static / dynamic gripping method based on binocular vision is studied.The design realization includes: MFC image acquisition based on VC6.0 and manipulator control interface development;image processing module based on the Opencv and the D-H modeling parameter setting module;motion control module based on DMC.In this paper,the static / dynamic autonomous discriminating and crawling system.is designed and developed for the REbot-V-6R 6-DOF manipulator based on the visual-assisted.The scoring process of static target and the way of dynamic target crawling are designed,and the crawling experiment is completed on the platform of manipulator to verify the rationality of target recognition,detection and tracking and manipulator kinematics model.