Research On Human-following Method Of Mobile Robot Based On Kalman Filter

With the development of computer technology and artificial intelligence,mobile robot has been widely used in the areas of industry,service industry,military,medicine and so on.Specially,the human-following technology is one of the important problems in the mobile robot technology.The human-following robot can follow the human target in real time.The key issue is how to improve the stability,accuracy and efficiency of the following system.This paper mainly studies the human-following method based on Kalman filter,aiming to improve the accuracy and robustness of the following system.The main work and results of the thesis are as follows:1.In view of the shortcomings of existing mobile robot platforms with poor scalability and low code reusability,a mobile robot platform based on ROS(Robot Operating System)is built,which provides corresponding software and hardware support for the implementation of the algorithm.2.For the problem of inaccuracy of human-following robot with ordinary linear Kalman filter,a following algorithm based on hypothesis Kalman filter is proposed.Compared with standard Kalman filter,this method reduces the following error of mobile robot and improves the stability of the human-following system.SVM(support vector machine)is used to detect and identify the human body for laser radar and DSST(Discriminative Scale Space Tracker)is used to realize the tracking of human body for RGB-D cameras.3.For the problem of low accuracy and poor stability of the human-following robot with a single sensor,the CI fusion method is used to fuse the information of the laser radar and RGB-D camera to achieve the follow-up of the human target.Compared with a single sensor,this method has better robustness and higher precision.4.The TurtleBot robot and OptiTrack positioning system are used as the experimental platform.The software system is developed under the environment of ROS and linux.Then,the relevant experiments are verified to show the effectiveness and robustness of the proposed following algorithm..