抄録
One of the important issues for mobile robot application is to solve online collision-free trajectory for navigation. For this, it is necessary to monitor walking people and forecast their motions. The problem to track multiple people using walking model from a mobile robot in motion is investigated in this paper. And real-time path planning methodology for a mobile robot using motion data from the people tracking is also studied. An experimental test-bed which consists of a mobile robot 'Yamabico' and a LRF sensor is employed in some experiments to track multiple walking people. Geometric characteristics around the legs of walking humans are analysed to detect their position from the data from LRF sensor. And EKF algorithm with walking model and walking frequency tracker proposed in this paper is utilized to track and anticipate the position of humans. The effectiveness of the proposed method is also evaluated some experiments for tracking multiple walking humans and path planning in indoor environment.
