Human Body Attitude Recognition And Robot Control Based On Kinect

With the rapid development of robot technology,the application of robots in industry has become a sign of the maturity of industrial technology.In recent years,robot control technology,and the way of interaction between people and robots become more and more simple,Interactive way gradually transferred from traditional teaching control to a more natural way of interaction and somatosensory control.The robot can control the movement towards the robot through the human body.And use of human movements and gestures to control the robot is a good way to reduce a lot of tedious robot teaching and program operation,promoting it more convenient and fast.The key technology of human computer interaction are the recognition of human actions and gestures based on machine vision.Through the domestic and foreign research on the development of somatosensory technology,found that most of the current somatosensory technology or tend to the game entertainment industry,in the application of the robot is still relatively lacking.Kinect somatosensory device is a new human-computer interaction device,it can capture,track,identify the body's movements,gestures and voice.Kinect somatosensory devices can also get the body's color images,depth data and human skeletal images.This topic is by referring to Microsoft's Kinect somatosensory equipment to carry out the body's attitude recognition research and robot somatosensory control simulation to achieve.The identification of human posture is the key to the development of somatosensory technology.It is necessary to design the skeletal data stream information of the target object in the real-time reproduction of the skeletal image.Kinect somatosensory device to connect Visual Studio 2010 to build the development environment,the target object of the skeleton frame object and the joint data acquisition,access to the 20 human nodes to connect,so that the human posture can be real-time reproduction of bone images come out.The robot 's somatosensory control simulation part is used to analyze the kinematics of the robot' s motion trajectory and joint angle.The angle of the robot is transformed into thejoint angle of the robot by using the angle measurement algorithm.And then by the Kinect somatosensory device to obtain the current scene of the 3D point cloud data.The point cloud data information is obtained by filtering the outliers and the redundant points in the main point cloud in the point cloud data.And then through the operator's finger target point coordinate data information,and then converted into the robot model of the joint angle of the data,in order to achieve control of robot movement.