Skip to Main Content
Nowadays, human-following is becoming one of the key issues of a human-robot interaction. An LRF (Laser Range Finder) might be the appropriate sensor in dealing with a human-following since it gives accurate information on a distance. Both extracting leg images from an LRF and interpreting those images over a time period as a same person are important. In this paper, we propose how to group legs into a person from laser scanned images and how to track a target person robustly over a time period using attributes we defined. The proposed methods were successfully verified through an experiment.