Aerial Maritime Vessel Detection and Identification

2025-07-09

AI Summary

Autonomous visual identification of maritime vessels under GNSS-denied conditions remains challenging—targets are described only visually, lack prior positional information, and require large-scale search under stringent computational constraints of embedded platforms. Method: This paper proposes an end-to-end vision-driven framework integrating ship detection, recognition, and localization. It employs a lightweight YOLOv8 for real-time detection, combines SIFT feature matching with hue histogram distance for fine-grained visual discrimination, and leverages multi-view geometric triangulation for GNSS-free localization. Contribution/Results: Evaluated on the MBZIRC 2023 real-world competition dataset, the system demonstrates robustness to complex viewpoint variations, achieves detection and recognition accuracy approaching the Oracle benchmark, and maintains bounded localization error. The framework significantly enhances autonomy and practicality for satellite-independent maritime surveillance, enabling reliable operation on resource-constrained embedded hardware.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Data transparency and provenance

Abstract

Autonomous maritime surveillance and target vessel identification in environments where Global Navigation Satellite Systems (GNSS) are not available is critical for a number of applications such as search and rescue and threat detection. When the target vessel is only described by visual cues and its last known position is not available, unmanned aerial vehicles (UAVs) must rely solely on on-board vision to scan a large search area under strict computational constraints. To address this challenge, we leverage the YOLOv8 object detection model to detect all vessels in the field of view. We then apply feature matching and hue histogram distance analysis to determine whether any detected vessel corresponds to the target. When found, we localize the target using simple geometric principles. We demonstrate the proposed method in real-world experiments during the MBZIRC2023 competition, integrated into a fully autonomous system with GNSS-denied navigation. We also evaluate the impact of perspective on detection accuracy and localization precision and compare it with the oracle approach.

Problem

• Detect vessels without GNSS using UAV vision
• Identify target vessels via visual cues alone
• Localize targets under computational constraints

Innovation

• Uses YOLOv8 for vessel detection
• Applies feature matching and hue analysis
• Localizes target with geometric principles