Vision-Based Monte Carlo Localization without Measurement: A Qualitative Approach during Update Phase

This paper presents a qualitative approach for updating the particles used by Monte Carlo Localization (MCL) during a mobile robot localization procedure. The combination between MCL and qualitative data will be called, in this article, Hybrid Localization. The motivation of using qualitative data is to obtain a level of abstraction closer to the human categorization of space and, also, to have a more effective way of interaction between robots and humans. The proposal uses the concept of a qualitative ego sphere, whereby the robot will perceive the world using qualitative relations. As RoboCup Humanoid League offers a challenging domain for robot localization, this environment was used to perform the experiments of this work, where experiments consisted of comparing the robustness of the proposed approach to a traditional vision based MCL model. The results allowed us to conclude that the use of qualitative data can show similar performance when compared to the traditional Vision Based MCL, bringing the advantage of being closer to the way humans reason about space, which can improve the communication between robots, as well as the development of high-level strategies during a game.