Robot Localization Based on Visual Landmarks

Abstract

In this paper, we will consider the localization problem of the autonomous minirobot Khepera II in a known environment. Mobile robots must be able to determine their own position to operate successfully in any environments. Our system combines odometry and a 2-D vision sensor to determine the position of the robot based on a new triangulation algorithm. The new system uses different colored cylinder landmarks which are positioned at the corners of the environment. The main aim is to analyze the accuracy and the robustness in case of noisy data and to obtain an accurate method to estimate the robot's position.