Person Following with Dynamic Obstacle Avoidance

Person following is essential for many human-robot interaction applications. However, many existing person-following methods do not consider obstacles that have a serious impact on the person-following. In this paper, we focus on the obstacle avoidance problem in the person-following process and propose a dynamic obstacle avoidance path planning method. Our method integrated trajectory prediction into global and local planning to follow a person and avoid obstacles. The proposed method is applied to a practical robot person-following task as an experiment. The experimental results show that the proposed method can achieve the target person following robustly in the case of obstacles.