Bird's–Eye View Image for the Localization of a Mobile Robot by Means of Trilateration

In this paper we introduce a method for the localization of a low–cost mobile robot, based on the use of a monocular camera. We consider a robot moving in a bi–dimensional environment, where some landmarks are placed in known positions. The acquired image of the environment is converted into a bird's–eye view image, used to measure the distance of the robot from the landmarks, to compute the robot's position by means of trilateration. The proposed strategy is able to compute both the position and the orientation of the robot.