Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Recovering the 3D UUV Position using UAV Imagery in Shallow-Water Environments

Antun Đuraš, Matija Sukno, Ivana Palunko

TL;DR

This work tackles underwater 3D UUV localization in shallow waters without relying on acoustic ground-truth systems. It proposes a vision-based pipeline that fuses UAV altitude, UUV depth, and 2D UUV image coordinates to recover the UUV’s 3D position in the camera frame and then convert it to GPS coordinates. The approach is validated both in simulation and through field experiments, achieving sub-meter accuracy and enabling geo-referenced data collection with low cost. The results demonstrate viability as a practical alternative to acoustic methods for shallow-water localization and dataset annotation, while highlighting areas for improvement in tracking robustness and surface refraction modeling.

Abstract

In this paper we propose a novel approach aimed at recovering the 3D position of an UUV from UAV imagery in shallow-water environments. Through combination of UAV and UUV measurements, we show that our method can be utilized as an accurate and cost-effective alternative when compared to acoustic sensing methods, typically required to obtain ground truth information in underwater localization problems. Furthermore, our approach allows for a seamless conversion to geo-referenced coordinates which can be utilized for navigation purposes. To validate our method, we present the results with data collected through a simulation environment and field experiments, demonstrating the ability to successfully recover the UUV position with sub-meter accuracy.

Recovering the 3D UUV Position using UAV Imagery in Shallow-Water Environments

TL;DR

This work tackles underwater 3D UUV localization in shallow waters without relying on acoustic ground-truth systems. It proposes a vision-based pipeline that fuses UAV altitude, UUV depth, and 2D UUV image coordinates to recover the UUV’s 3D position in the camera frame and then convert it to GPS coordinates. The approach is validated both in simulation and through field experiments, achieving sub-meter accuracy and enabling geo-referenced data collection with low cost. The results demonstrate viability as a practical alternative to acoustic methods for shallow-water localization and dataset annotation, while highlighting areas for improvement in tracking robustness and surface refraction modeling.

Abstract

In this paper we propose a novel approach aimed at recovering the 3D position of an UUV from UAV imagery in shallow-water environments. Through combination of UAV and UUV measurements, we show that our method can be utilized as an accurate and cost-effective alternative when compared to acoustic sensing methods, typically required to obtain ground truth information in underwater localization problems. Furthermore, our approach allows for a seamless conversion to geo-referenced coordinates which can be utilized for navigation purposes. To validate our method, we present the results with data collected through a simulation environment and field experiments, demonstrating the ability to successfully recover the UUV position with sub-meter accuracy.
Paper Structure (14 sections, 27 equations, 9 figures, 1 table)

This paper contains 14 sections, 27 equations, 9 figures, 1 table.

Table of Contents

  1. Introduction
  2. Problem statement
  3. Related Work
  4. Methodology
  5. The pinhole camera model
  6. Line-Plane Intersection
  7. Recovering the 3D UUV position
  8. Conversion to GPS coordinates
  9. Results
  10. Simulation Environment
  11. Data Acquisition Setup
  12. Ground truth system
  13. Discussion of results
  14. Conclusion and Further Research

Figures (9)

  • Figure 1: Illustration of the main components of our method, the UAV monitoring the operating area of the UUV
  • Figure 2: Image demonstrating the flickering effects of sunlight, often seen in shallow-water images captured by an UUV camera
  • Figure 3: System diagram illustrating sensor information flow from the robots (blue) to main processing components (yellow) used to obtain the UUV position estimate.
  • Figure 4: The pinhole camera model
  • Figure 5: Snapshot of the simulation environment
  • ...and 4 more figures