ArUco Maker based localization and Node graph approach to mapping

This paper explores a method of localization and navigation of indoor mobile robots using a node graph of landmarks that are based on fiducial markers. The use of ArUco markers and their 2D orientation with respect to the camera of the robot and the distance to the markers from the camera is used to calculate the relative position of the robot as well as the relative positions of other markers. The proposed method combines aspects of beacon-based navigation and Simultaneous Localization and Mapping based navigation. The implementation of this method uses a depth camera to obtain the distance to the marker. After calculating the required orientation of the marker, it relies on odometry calculations for tracking the position after localization with respect to the marker. Using the odometry and the relative position of one marker, the robot is then localized with respect to another marker. The relative positions and orientation of the two markers are then calculated. The markers are represented as nodes and the relative distances and orientations are represented as edges connecting the nodes and a node graph can be generated that represents a map for the robot. The method was tested on a wheeled humanoid robot with the objective of having it autonomously navigate to a charging station inside a room. This objective was successfully achieved and the limitations and future improvements are briefly discussed.