Abstract
This paper describes a person recognition method using IMU and LRF by a mobile service robot. The robot is equipped with two LRFs, and the IMU is attached to a leg of the target person. Walking conditions, such as foot strike timing and stopping/turning period, are measured by the LRF and the IMU independently. We compare the walking conditions measured by the LRF and the IMU and calculate the degree of difference between these conditions. For a robust target tracking, we identify the target person using a Baysian estimation. In the experiment, the robot recognizes the target person successfully even if the person is occluded by the other persons.
