Dual-Type Marker Fusion-Based Underwater Visual Localization for Autonomous Docking
Chunyang Zhao, Huijie Dong, Jian Wang, Tiezhu Qiao, Junzhi Yu, Jieyu Ren
Abstract
Underwater localization is necessary for autonomous operation of underwater robots. Limited field of view for onboard vision systems can result in poor reliability of visual localization for underwater robots during autonomous docking. This article proposes a dual-type marker fusion-based visual localization method for underwater autonomous docking. First, a visual localization scheme integrating light sources and ArUco markers is designed according to positioning requirements based on varying fields of view for onboard vision in autonomous docking operations. The arrangements of the light sources and ArUco markers are both analyzed and optimized to ensure the reliability and efficiency of visual localization. Then, an underwater visual localization algorithm with dual marker fusion is proposed for high-accuracy localization with a restricted field of view. Extended underwater experiments are conducted to verify the feasibility and stability of the proposed method across various distances. The obtained results provide valuable references for the design and optimization of visual navigation systems for underwater robots.