부산대학교 도서관

Fiducial marker (FM) systems have a wide range of applications, such as augmented reality, human-robot interaction, medical surgery, robot navigation, and swarm robotics. Moreover, FMs enable to calibrate optical sensors, which is a crucial step in most computer vision algorithms. A variety of considerations should be taken into account to decide which FM system is suitable for a particular application. This paper proposes a list of recommended criteria that allow to compare FM systems. The criteria support a FM ranking according to their properties and enable users to choose a set of FMs best suited for a specific task. A number of the recommended criteria efficiency and performance were validated in the Gazebo simulation and in a real world environment using Servosila Engineer mobile robot.