Topological localization with kidnap recovery using sonar grid map matching in a home environment

This paper presents a method of topological localization with kidnap recovery capability in a home environment using only low-cost sonar sensors. The proposed method considers both pose tracking and relocation problems. The pose tracking is achieved by calculating node probability using grid-map matching and relative motion model. The relocation method detects the kidnap automatically and recovers it using multiple hypothesis tracking. After kidnap recovery, it also provides a criterion for selecting a reasonable hypothesis for returning to the pose tracking stage autonomously. Experimental results in a real home environment verify that the proposed localization method provides a reliable and convergent node probability when the robot is kidnapped.

► We propose a topological localization method based on the grid-map matching.
► The proposed method consists of pose tracking stage and relocation stage.
► The pose tracking method provides a convergent node probability through Bayesian update process.
► The relocation method detects and recovers kidnap automatically.
► The performance of the proposed method was verified by experiments in a real home environment.