A Visual Servoing Scheme for Autonomous Aquaculture Net Pens Inspection Using ROV

Area of Science:

• Marine Engineering
• Robotics
• Aquaculture Technology

Background:

• Aquaculture net pen inspection is crucial for maintaining fish health and farm stability.
• Underwater inspection is challenging due to dynamic and complex environments.
• Remotely Operated Vehicles (ROVs) provide a viable solution for autonomous underwater monitoring.

Purpose of the Study:

• To integrate an ROV with a visual servoing scheme for automated inspection and tracking of aquaculture net pens.
• To develop and validate a vision-based positioning system for ROV navigation along net pens.

Main Methods:

• A modular vision-based positioning scheme was developed, including object detection, pose generation, and closed-loop control.
• Traditional computer vision methods were used to process images of identifiable ropes on the net.
• A vision triangulation method determined the ROV's position relative to the net plane.
• A closed-loop control law guided the ROV's traversal along the net for inspection.

Main Results:

• The proposed vision-based scheme was successfully implemented and tested via simulations and field experiments.
• The system demonstrated satisfactory performance in inspecting and tracking aquaculture net pens.
• The ROV system effectively traversed from top to bottom along the net plane.

Conclusions:

• The integrated ROV and visual servoing system offers a reliable and effective method for aquaculture net pen inspection.
• This technology can supplement existing traditional inspection methods, enhancing efficiency and safety.
• The system contributes to improved fish farm management through automated monitoring.