Research On Vision And Behavior Of NAO Robot

Humanold Robot has always been an important research for years and theory of Robot's visual theory and control technology are the Theoretical basis and key techniques for realizing tis autokinetic movement.Selecting Robo Cup Standard Platform League as the research object, this report emphasizes the Theprectical and experimental research of the NAO Robot's visual theory and control technology.The contents and achievements in this research mainly include:(1) Based on the visual feature recognition of the robot, the report discusses in detail about image acquisition, image processing, targeting and positioning of the robot.We finally adopt HSV color spaces through experimental contrast, proposing a more effective way of location algorithm based on the unary Scaling algorithm.(2) Studying on the robot's kinesiology, setting up the Connecting rods model, completing analysis of inverse and direct kinematics of the robot by D-H Law, we successfully accomplishing its grabbing gestures with its right arm. We propose an optimization by comparing different grabbing tracks.Using NAO robot to complete the experimental task of picking up trash,verify the correctness and effectiveness of the improved algorithm for robot vision and control technology.According to the deficiencies of experiments,we proposed experiental optimization algorithm and modifications.