TURBEAMS
3D turbidity mapping using multibeam sonar and deep learning
TURBEAMS: 3D Turbidity Mapping
TURBEAMS (Towards 3D turbidity by correlating multibeam sonar and in-situ sensor data) is a national research project aimed at developing innovative methods for 3D turbidity and suspended particulate matter (SPM) imaging in aquatic environments.
Project Overview
This BELSPO-funded project (New RV Belgica program) combines multibeam sonar technology with machine learning to create high-resolution 3D maps of water quality parameters.
My Contribution
During my postdoctoral research at Ghent University (2023), I contributed to the project by:
- Deep Learning Model Development: Built neural networks linking multibeam water-column backscatter with turbidity/SPM measurements
- Data Processing Optimization: Scaled data handling to process millions of acoustic data points per survey session
- Python Workflow Development: Created efficient data flows aligned with the project’s 3D imaging objectives
Technical Innovation
The project represents a significant advancement in:
- Acoustic-Optical Data Fusion: Combining multibeam sonar with in-situ optical sensors
- Real-time Processing: Developing workflows for near real-time turbidity mapping
- 3D Visualization: Creating volumetric representations of water quality parameters
Impact
This work supports:
- Environmental Monitoring: Enhanced capabilities for tracking sediment transport and water quality
- Marine Science: New tools for understanding aquatic ecosystem dynamics
- Maritime Operations: Improved navigation and dredging support
Collaboration
The project involves multiple Belgian institutions and represents a collaborative effort to advance marine technology and environmental monitoring capabilities.
Project Page: BELSPO TURBEAMS
TURBEAMS demonstrates the power of combining advanced acoustic sensing with machine learning for environmental monitoring applications.