Evaluation of Visual Markers Detection Used for Autonomous Mobile Robot Docking Navigation

The paper deals with the evaluation of visual markers detection. The markers are used for the navigation during the autonomous mobile robot docking procedure. In order to navigate to the docking station, the station must be easily detectable. AR (augmented reality) markers are designed with respect to the precision and low computational load of detection. The paper evaluates the properties of detection for AR markers supported in ROS, in particular the Alvar and Apriltag markers detectors. Experiments handle both single and double markers, precision is evaluated with respect to the distance, angle and relative tangential velocity of robot’s camera towards the marker.