Vision-Based Bubble Measurement System for Overlapping Air Bubbles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current air bubble detection and measurement systems are inaccurate, labor-intensive, and unable to consistently measure bubble size and velocity, particularly for small bubbles and those that overlap or split, leading to incomplete and inaccurate data.
Innovation Solution
A vision-based bubble measurement system using an imaging device and controller with a neural network algorithm for image processing, including a pairing module to compare successive images and perform binary image classification and region-based convolutional neural network analysis to accurately measure and track air bubbles, even in cases of overlap or split bubbles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If vision-based measurement systems are used to image bubbles, then bubble dimension can be measured, but detection accuracy deteriorates for small bubbles and overlapping bubbles
Solution Approach 1:
The system divides the bubble measurement process into multiple sequential imaging stages. Bubbles are imaged at different positions along the flow path, allowing the system to track individual bubbles through multiple frames. This segmentation of the measurement process enables accurate tracking of small bubbles and proper handling of overlapping bubbles by observing their movement and separation across time steps.
Solution Approach 2:
The system performs preliminary imaging of bubbles at an first position before they reach the detection point. By capturing bubble images early in the flow path and tracking them through subsequent positions, the system establishes baseline measurements and identifies individual bubbles before they may overlap or become difficult to distinguish, improving detection accuracy for small and overlapping bubbles.
2Quantity of substance
If aggregated collection of many air bubbles is used for calibration, then average bubble size can be determined, but consistency of individual bubble size cannot be validated
Solution Approach 1:
The system replaces the manual aggregated collection method with an automated vision-based measurement system. The imaging device captures individual bubble images, and the controller automatically processes these images to measure each bubble's dimensions. This substitution enables precise measurement of individual bubble sizes and their consistency, eliminating the limitations of manual aggregation methods.
Solution Approach 2:
The system performs self-measurement by automatically capturing, processing, and analyzing bubble images without requiring manual intervention. The controller autonomously identifies individual bubbles in the images, measures their dimensions, and calculates statistical parameters including size consistency. This self-service capability provides both total volume and individual bubble size validation simultaneously.
3Reliability
If known calibration systems are used, then bubble generator can be tested, but the process becomes labor intensive and time consuming
Solution Approach 1:
The system replaces labor-intensive manual calibration procedures with automated vision-based measurement. The imaging device continuously captures bubble images, and the controller automatically processes these images to validate bubble generator performance. This automation eliminates manual measurement steps, significantly reducing calibration time while maintaining or improving validation reliability.
Solution Approach 2:
The system enables continuous measurement and validation of bubble generator performance by maintaining continuous imaging and processing. Rather than discrete manual measurements, the automated system continuously tracks bubbles through the flow path, providing ongoing validation data. This continuous operation reduces total calibration time while improving the reliability of bubble generator testing.
4Device complexity
If dimension measurement of only the first bubble in the image is recorded, then processing is simplified, but dimension estimation accuracy deteriorates
Solution Approach 1:
The system segments the bubble measurement process into multiple imaging positions along the flow path. By imaging bubbles at different positions and tracking their movement, the system can measure dimensions of multiple bubbles including overlapping ones. The controller identifies and tracks individual bubbles across frames, recording dimensions for each bubble rather than only the first one, improving dimension estimation accuracy without excessive complexity increase.
Data Source
AI summary
A bubble measurement system includes a bubble detector including a vessel having a flow path configured to receive a flow of fluid including air bubbles from a bubble generator and an imaging system. The imaging system includes an imaging device for imaging the fluid and air bubbles in the flow path of the vessel of the bubble detector. The imaging system has an imaging controller coupled to the imaging device and receiving images from the imaging device. The imaging controller processes the images to measure bubble size of each air bubble passing through the bubble detector. The imaging controller includes a pairing module comparing successive images and the air bubbles in successive images to measure all bubbles flowing through the vessel.