Research on multi-feature fusion-based terminal visual autonomous docking technology for AUV

Objectives To address the insufficient docking accuracy of autonomous underwater vehicles (AUVs) in complex underwater environments, a vision-guided method based on multi-feature fusion is proposed. Methods Leveraging a self-developed four-propeller rudderless vector propulsion AUV, the dark channel prior (DCP) dehazing algorithm was adopted for image enhancement. Improved Canny edge detection and color threshold segmentation were combined to achieve multi-feature fusion. The minimum enclosing circle method was utilized for circle center positioning, and coordinate system transformation was implemented to calculate relative position and orientation for docking. Results Unity 3D simulations and pool experiments demonstrated distance-dependent trends: both mean difference and root mean square error decreased as docking distance reduced. Closer distances yielded higher visual ranging accuracy and docking precision. When the docking distance was less than 2 m, the positioning error fell below 5 cm, with an overall success rate of 88%.Conclusions The proposed method meets AUV autonomous docking accuracy requirements and provides a highly robust solution for underwater equipment recovery.