抄録
This paper describes an unconstrained gait measurement method by using a mobile robot as one of a service for perceiving human's health. Human's gait is measured based on heel recognition by the mobile robot equipped with a camera. The mobile robot detects target person's heels from obtained images using particle filter, and calculates relative positions of heels compared with own odometry information to estimate strides and walking speed while tracking a target person. Experimental results show a usefulness of proposed method by comparison of estimated value with actual measured value.
