Research On Intelligent Recognition And Grasp System Of Robot Based On ROS

The ability of the robot to make the next decision through the surrounding environment reflects the level of intelligence.Because of working in an unstructured environment,the robot needs to be able to perceive the world.The cost of Kinect sensor equipment is very low that can provide real-time point cloud and two-dimensional images,and the bottom layer is based on structured light technology.PCL(Point Cloud Library)which use modular design ideas is the modern c++ library of 3D point cloud processing.From an arithmetic point of view,PCL is designed to combine a large number of 3D algorithms to handle a large number of point cloud data.EOD robots play an increasingly important role in the EOD task,the main work of the arm is to handle,destruction and other tasks.Because of the instability of the emotions,ability and the increase of the training cost,the operators have a higher demand for intelligent,so it is important significance to research intelligent recognition and grasping system of EOD robot.In this paper,the point cloud data model library of unknown explosives is constructed,and the main processing steps are integrated into the ROS of the robot operating system,such as segmentation,feature extraction and so on.The main research work is as follows:1.We study the method of segmentation of the original point cloud,and the basic filter.Then perform plane segmentation,find all the planes in the scene.By giving a convex hull computed from the plane coefficients,it will extrude it a certain height to create a prism,and then give back all points that lie inside,then split out the object.According to the characteristics of the feature operator in the PCL,we choose the VFH(Viewpoint Feature Histogram)descriptor to establish the object cloud model database.2.This paper uses PCL to perform image preprocessing of point cloud images acquired by real-time robot,including filtering,reconstruction,segmentation and feature extraction.It is integrated into ROS(Robot Operating System)to complete the Intelligent identification of of explosives,and the application of ICP algorithm for 6 degrees of freedom pose estimation.3.The 3D model of five-degree-of-freedom manipulator is established.The kinematics model of the manipulator was established by DH method,including the positive and inverse kinematics.We use the Matlab Robotics Toolbox to simulation.The mechanical hardware system was developed,including the power module,the main controller module,the motor control module,etc.Finally,the paper summarizes the full text,analyzes the problems existing in the present research,and puts forward some suggestions and prospects for further research.