A framework for the design of person following behaviors for social mobile robots

摘要

This paper presents an environment for design of interactive behaviors for social robots. Its structure consists of three levels which are: perception, decision and action. Their principles are briefly described and we detail more specifically implementation of people searching and monitoring in cluttered environments. We describe in particular the methods for people detection, state estimation and trajectory generation for interactive monitoring. We demonstrate experimentally that the designed reasoning system is able to regulate the robot behavior in complex situations. The robustness of the framework is evaluated experimentally for person following in his home space that may lead the system to lose the person detection because of occlusions and to use shortcuts.