Localization algorithm using virtual label for a mobile robot in indoor and outdoor environment

Localization for autonomous mobile robot is one of the most important factors. If current position of robot is unknown, it should be impossible to solve a problem such as how to decide and where to go. There are various localization methods developed by previous researches; localization using RFID tag, ultrasonic and image processing and recognizing landmark. However, these methods have problems and faults. In indoor, RFID, vision and ultrasonic sensors are only detected a restricted space and easily effected by height, interval and direction of installed sensors. In outdoor, it is difficult to detect current position because of disconnected signal and error of GPS which has used in the most of research. This paper proposes new localization algorithm which is assigned virtual label along with map building through information acquired by using a 2D laser range finder, an encoder and a GPS in indoor/outdoor environments. The proposed algorithm for virtual label is verified by applying a mobile robot both indoor and outdoor.