Opti-Acoustic Scene Reconstruction in Highly Turbid Underwater Environments

2025-08-05

AI Summary

In highly turbid underwater environments, severe optical degradation and low-resolution sonar data render conventional single-modal 3D reconstruction ineffective. To address this, we propose a real-time optical-sonar cross-modal scene reconstruction method. Our key innovation lies in replacing traditional feature-point matching with a robust region-of-interest (ROI)-based matching strategy that fuses monocular visual cues with sonar-derived range and elevation measurements. This approach effectively mitigates the dual challenges of image blur in turbid water and sparse sonar point clouds. Extensive evaluation across multiple turbidity levels and real-world harbor deployments demonstrates substantial performance gains over both vision-only and sonar-only baselines. The implementation is open-sourced to ensure reproducibility and foster community collaboration.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Multi-modal Vision
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Scene reconstruction is an essential capability for underwater robots navigating in close proximity to structures. Monocular vision-based reconstruction methods are unreliable in turbid waters and lack depth scale information. Sonars are robust to turbid water and non-uniform lighting conditions, however, they have low resolution and elevation ambiguity. This work proposes a real-time opti-acoustic scene reconstruction method that is specially optimized to work in turbid water. Our strategy avoids having to identify point features in visual data and instead identifies regions of interest in the data. We then match relevant regions in the image to corresponding sonar data. A reconstruction is obtained by leveraging range data from the sonar and elevation data from the camera image. Experimental comparisons against other vision-based and sonar-based approaches at varying turbidity levels, and field tests conducted in marina environments, validate the effectiveness of the proposed approach. We have made our code open-source to facilitate reproducibility and encourage community engagement.

Problem

• Reconstructing underwater scenes in highly turbid environments
• Overcoming limitations of monocular vision and sonar in turbid water
• Real-time opti-acoustic fusion for accurate 3D scene reconstruction

Innovation

• Combines sonar and camera for turbid water
• Matches image regions to sonar data
• Uses sonar range and camera elevation