Session: Government Agency Student Posters

Paper Number: 172888

Development of an Integrated Underwater Imaging System for Real-Time Bubble Characterization and Flow Measurement 

Bubble ebullition is ubiquitous in aquatic environments and contributes significantly to methane emissions in ocean and inland waters, including rivers and lakes, continental margins, and the Arctic, representing a critical component of global greenhouse gas cycling. Understanding the bubble size distribution, bubble rise velocity, and the turbulent flows agitated by rising bubbles is crucial for accurately estimating methane flux into the water column and the atmosphere in shallow waters. Current measurement techniques, however, cannot characterize plume structure and bubble wake turbulence near the source. Currently, no equipment exists that can quantify both bubble dynamics and flow dynamics at the same time to illustrate the fine-scale structure in submarine seeps and bubble ebullition inland water bodies. We address this limitation by developing and validating a novel underwater bubble imaging system that enhances aquatic environmental monitoring by combining particle imaging velocimetry (PIV) flow measurement with real-time bubble size detection, enabling better understanding of bubble leakage mechanisms.

We developed a dual-configuration imaging system that combines traditional PIV with artificial intelligence-driven bubble detection for comprehensive underwater flow characterization. The system utilizes an NVIDIA Jetson embedded computing platform, enabling real-time processing capabilities not previously available in underwater applications. The hardware components include a laser light source for PIV measurements and an Ethernet monochrome camera for image capture, with an LED white pad positioned behind the measurement volume for enhanced bubble size detection. This integrated design eliminates the need for separate systems that conventional methods require for flow and bubble measurements.

Two distinct operational modes were implemented: the PIV measurement configuration incorporates an embedded control system to perform in-situ flow field measurements, while the real-time detection configuration employs a custom-trained bubble detection model to measure bubble size distribution and provide statistical information in real time. We trained the embedded AI detection model using 5,000 synthetic bubble plume images to achieve robust detection across varying lighting conditions and bubble morphologies.

Laboratory validation was conducted under controlled conditions. For the PIV measurement configuration, we tested the prototype with a single-phase jet flow, bubble-in-chain and bubble plume, comparing results with measurements from a conventional PIV system to establish accuracy benchmarks. We systematically analyzed bubble rise velocity and the complex flow patterns surrounding the bubbles, including turbulent intensity and vorticity fields. For the real-time detection configuration, we validated the AI detection model by comparing bubble sizes identified by the algorithm with those calculated from multiple-angle camera images, conducting the system statistical error analysis to validate the accuracy of our method.

Results demonstrate excellent agreement in bubble size measurements between our system and reference methods, with less than 8% error in diameter measurements across bubble sizes ranging from 1-10mm. Flow velocity measurements showed good agreement with conventional lab PIV systems. The real-time detection configuration successfully processed bubble images at 2 frame-per-second (fps), providing instantaneous size distribution analysis with processing latency under 200 ms. The system's ability to capture bubble dynamics and surrounding flow fields offers substantial improvements in underwater measurement capabilities.

These findings validate the system's potential for field deployment in natural aquatic environments, where traditional laboratory-based measurement techniques are impractical. The compact, submersible design enables deployment in various water environments of methane seepage areas. Applications for this technology include studies of bubble ebullition in natural water bodies, lake oxygenation monitoring, wastewater treatment optimization, enhanced oil recovery processes, and other bubble-driven industrial applications.

Presenting Author: Xuchen Ying University of Missouri - Columbia

Presenting Author Biography: Xuchen Ying is a Ph.D. student in Civil and Environmental Engineering at the University of Missouri–Columbia. He holds both bachelor’s and master’s degrees in Engineering Thermophysics from the University of Shanghai for Science and Technology (USST) in Shanghai. His research focuses on developing in situ bubble-plume measurement methods aimed at obtaining accurate, high-value physical statistics to support modeling and theoretical predictions. His work spans a range of methodologies, including remote-sensor system design, embedded-device development, and computer-vision–based data processing. His research has been published in Applied Energy, Physics of Fluids, Limnology and Oceanography, and other journals.

Authors:

Xuchen Ying University of Missouri - Columbia
Binbin Wang University of Missouri - Columbia