This paper presents a navigation method for an indoor mobile robot. The proposed method employs indoor lights as landmarks and an encoder for localization and ultrasonic sensors for obstacle avoidance.
Indoors lights are easily found in any man-made environment: for example, office, museum, exhibition hall, and so on. Therefore, if indoor lights are used as landmarks for mobile robot navigation, no extra setup is needed. In addition, since they emit lights, they have distinct grey level from the other objects in an image. Accordingly, indoor ligths as landmarks have an advantage that they can easily be extracted from the other objects by a simple binarization.
In this work, we implement the robot system with a single camera and ultrasonic sensors. The vision system calculates current positions of the robot from landmarks and ultrasonic sensors are used to avoid obstacles.
Localization experiment by the vision system is performed to show the plausibility of the proposed method using indoor lights as landmarks. The accuracy of the localization is presented and real-time issue is also considered. To evaluate the performance of the proposed method, robot navigation is performed by integrating the localization. method and obstacle avoidance by vector-field approach using ultrasonic sensors.