Research On Target Recognition And Path Planning Based On Vision For Soccer Robot

Robot soccer competition is a product of the combination of football and robotics,artificial intelligence,behavioral decision-making and other high-tech,and is a good practice platform.This paper takes small group soccer robot as the research object,mainly studies the calibration of visual camera,threshold calibration of color image segmentation,calibration of robot field,determination of robot pose and robot path planning.The main research contents of this paper are as follows:Firstly,in order to improve the calibration accuracy of the camera,a method of using the average pixel error to measure the standard accuracy is proposed.The calibration of the camera is completed by Zhang Zhengyou.Aiming at the problem that illumination affects the threshold of color image segmentation,a threshold component calibration method is proposed and a comparative experiment is carried out.Secondly,the mathematical model of image coordinate system and site coordinate system transformation is established.The mathematical model is modified for the deflection problem between the camera and the field during the installation process.A field calibration method based on shape template matching was proposed and the calibration experiment was carried out on the site.Thirdly,the mathematical model of pose description for the robot color scale is established,and the difference between the actual position caused by the height of the robot and the target recognition position is compensated.The position,direction angle and label of the robot are identified..Finally,the common methods of robot path planning are studied.Based on the artificial potential field method,a step evaluation function is established for the problem of convergence speed.The force of the two adjacent steps of the robot and the direction angle of the robot are introduced into the step.The evaluation function was long and a comparative test was conducted.