AUV Docking and Positioning Technology based on Apriltag and Machine Vision

As a powerful tool for dynamic ocean observation, the Autonomous Underwater Vehicle (AUV) needs to dock with a base station to ensure long-term stability in ocean monitoring. Docking and positioning are crucial for autonomous underwater docking. This paper proposes an AUV docking and positioning technique based on AprilTags and machine vision. In this method, multiple AprilTag serves as passive markers for short-range visual guidance. By integrating machine vision, this approach provides highly accurate and stable visual positioning information during the docking navigation phase. Even when some AprilTags are obstructed at extremely short distances between the AUV and the station, this technique can still calculate pose information, thereby enhancing the success rate of autonomous underwater docking. To validate the positioning results, a series of simulations and underwater experiments were conducted, confirming the accuracy and effectiveness of the positioning algorithm.