Object-based Place Recognition for Mobile Robots using Laplace's Rule of Succession

Abstract

Place recognition estimating a robot's location from various sensor data is one of the most active research field in robotics, as it is a prerequisite step for robots to perform more intelligent and complex tasks. In this paper, we present an object-based place recognition method using objects as a key intermediate representation of places. Objects are detected by a mobile robot equipped with a camera, and object map is constructed to contain object information of places. Then, locations are recognized by object information located at places. As the proposed method considers not only the type of the object but also the number of occurrences of objects, it is possible to distinguish places that could not be distinguished from the existing method. Simulation and experimental results are presented to verify the proposed method.