BubbleID: A Deep Learning Framework for Bubble Interface Dynamics Analysis

TL;DR

BubbleID is a deep learning framework that accurately analyzes bubble interface dynamics in boiling images, combining segmentation and tracking to study bubble behavior under various conditions.

Contribution

Introduces BubbleID, a novel deep learning architecture integrating Mask R-CNN and SORT for comprehensive bubble analysis in boiling images.

Findings

Effective in identifying static and dynamic bubble attributes.

Capable of analyzing bubble behavior across different heater surfaces.

Provides insights into bubble dynamics before and after critical heat flux.

Abstract

This paper presents BubbleID, a sophisticated deep learning architecture designed to comprehensively identify both static and dynamic attributes of bubbles within sequences of boiling images. By amalgamating segmentation powered by Mask R-CNN with SORT-based tracking techniques, the framework is capable of analyzing each bubble's location, dimensions, interface shape, and velocity over its lifetime, and capturing dynamic events such as bubble departure. BubbleID is trained and tested on boiling images across diverse heater surfaces and operational settings. This paper also offers a comparative analysis of bubble interface dynamics prior to and post-critical heat flux (CHF) conditions.