Human Target Following of Mobile Robot based on Lidar

Aiming at the problem of human body target following of mobile robot in indoor environment, a target following method of mobile robot based on lidar is proposed. In this method, lidar is used as the environment sensing sensor to obtain the laser scanning data of the environment in real time. The geometric features of human legs are defined by the arc features of human legs, and the support vector machine (SVM) method is used to train the sampling data set and predict the position information (tracking deflection angle and tracking distance) of the target human legs relative to the robot. Considering that the orientation of people may change greatly at each time in the process of movement, a hypothetical Kalman filter algorithm is proposed to predict and update the relative position relationship between robot and people, so that the robot can follow the target person more smoothly. The algorithm is tested on dfrobot robot, and the experimental results verify the effectiveness of the algorithm.