Enhanced YOLOv8 Ship Detection for Unmanned Surface Vehicles
Author Information
Author(s): Haijoub Abdelilah, Hatim Anas, Guerrero-Gonzalez Antonio, Arioua Mounir, Chougdali Khalid
Primary Institution: Universidad Politécnica de Cartagena
Hypothesis
Can an enhanced version of YOLOv8 improve ship detection accuracy and energy efficiency for unmanned surface vehicles in maritime surveillance?
Conclusion
The study demonstrates that the enhanced YOLOv8 model achieves a mean Average Precision of 0.99 and operates at 17.99 FPS with an energy consumption of just 5.61 joules.
Supporting Evidence
- The enhanced YOLOv8 model achieved a mean Average Precision of 0.99.
- The system operates at a speed of 17.99 FPS.
- Energy consumption was measured at just 5.61 joules.
Takeaway
This study shows how a smart computer program can help boats see and track other boats better while using less energy.
Methodology
The study developed a maritime surveillance system using an enhanced YOLOv8 model deployed on the NVIDIA Jetson TX2, focusing on real-time detection and tracking of ships.
Limitations
The study does not specify limitations, but potential challenges include the complexity of integrating advanced models and energy consumption during high-performance tasks.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website