Underwater vehicle tracks moving targets using improved depth and control methods
Underwater Visual Target Tracking with Target-Specific Depth Estimation and Adaptive Model-Fusion Predictive Control
Robotics
Summary
Tracking objects underwater is hard because cameras can’t easily guess how far away things are, and the objects can move unpredictably. The authors built a system for underwater robots that uses two cameras to better figure out how far the target is and where it is in 3D space. Their system also uses smarter control methods to steer the robot more smoothly toward the target by predicting how the target might move. Tests show their method works better than earlier systems in both simulations and real water.
What this means in practice
- •For autonomous underwater vehicle operators: Use the system to improve tracking of moving targets in underwater exploration or inspection tasks where depth sensing is unreliable.
- •For marine robotics developers: Integrate the proposed adaptive control and depth estimation methods into AUV navigation software for enhanced real-time target following.
Authors
Yuheng Zhou, Haiyang Cheng, Yanqi Feng, Pangkit Fong, Mei Xuan Lee, Marcus Gee, Chongrong Fang, Jianping He
Abstract
Vision-based underwater target tracking is challenged by unreliable depth measurements and unknown target motion. This paper proposes a stereo visual-servoing framework for an autonomous underwater vehicle (AUV). For perception, the framework derives a stable 3D relative state from stereo images through target-specific depth extraction and Kalman filtering. It constructs a target-depth mask from color, disparity, and temporal cues to select reliable target pixels, and then filters the resulting depth measurement and detected image center separately. For control, the framework decouples yaw regulation from translational control, avoiding computationally expensive coupled multi-DOF optimization and enabling real-time translational MPC. The translational controller employs adaptive model-fusion predictive control, combining constant-velocity and zero-velocity target models to accommodate different target-motion patterns. It updates the model weights using historical prediction errors and computes translational commands subject to actuation, following-distance, and field-of-view constraints. Through simulations and real-world experiments, we validate the effectiveness of the proposed framework and show it has better performance than existing frameworks.