Autonomous underwater vehicle docking system for energy and data transmission in cabled ocean observatory networks

Cabled ocean observatory networks (COON) are used for long-term all-weather observation of submarine scientific data, which contribute to low-carbon energy research. Autonomous underwater vehicles (AUV) with clean can provide active search capabilities by connecting the docking station (DS) on COON complete and data transmission in detection tasks. The AUV is guided optical landmarks a vision system short-range docking. In this study, we propose an tracking framework solve problem detecting failure caused incomplete landmarks. First, two-stage algorithm based CNN estimate 3D relative position orientation between DS during docking, including detect phase PnP pose estimator. Then extended Kalman filter Hungarian matching introduced improve robustness algorithm. reliability vision-based verified pool, field environment shown lake experiment combined long-range guidance. experimental results indicate that effectively leverage information enhance scope visual guidance