Localization Method for Autonomous Mobile Robot using Visual Features in Environment

For reliable navigation, a mobile robot needs to be able to recognize where it is in the world. The aim of this paper is to present a robust vision-based approach to mobile robot localization, which employ visual features in environment in the system. In our approach, the robot does not have to measure its own position precisely and accurately, and rely only to the visual features which could be extracted from the environment. The features extraction method works simply and leads to a robust localization system. The proposed localization method has been implemented and tested in both indoor and outdoor environments. We present several experiments demonstrating the reliability and robustness of our approach. The achieved results are presented and discussed.