IP Library Granted Patent US 11,995,860
Granted Patent B2
US 11,995,860 · App. 17/646,916 · Granted May 28, 2024

Bubble measurement system and method

Inventors: Shu Wang (Middletown, PA); Du Wen (Reading, PA); Roberto Francisco-Yi Lu (Bellevue, WA); Jiankun Zhou (Middletown, PA)
Assignee: TE Connectivity Solutions GmbH
G06T7/62G01B11/043G01B11/10G01B11/28G01F1/74G01N21/85G01N2021/1765G01N2201/12723G06T2207/20084
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,995,860
App. No.
17/646,916
Granted
May 28, 2024
Kind
B2
Abstract

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.

Claims (34)

1. A bubble measurement system comprising:

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 having an imaging device for imaging the fluid and air bubbles in the flow path of the vessel of the bubble detector, the imaging system having an imaging controller coupled to the imaging device and receiving images from the imaging device, the imaging controller processing the images to measure bubble size of each air bubble passing through the bubble detector, the imaging controller including a pairing module comparing successive images and the air bubbles in successive images to measure all bubbles flowing through the vessel, and the pairing module identifying bubble overlap conditions.

2. The bubble measurement system of claim 1 , wherein the pairing module compares successive images to determine when the air bubbles enter a field of view of the imaging device and when the air bubbles exit the field of view of the imaging device.

3. The bubble measurement system of claim 1 , wherein the pairing module includes a neural network algorithm to match air bubbles in the successive images.

4. The bubble measurement system of claim 1 , wherein the pairing module uses positional data of the air bubbles in the successive images and uses measurement data of the air bubbles in the successive images to match the air bubbles between the successive images.

5. The bubble measurement system of claim 4 , wherein the measurement data includes elongation data, perimeter data, and area data of the air bubbles in the successive images to match the air bubbles between the successive images.

6. The bubble measurement system of claim 1 , wherein the pairing module uses a Gale-Shapley algorithm to find optimal pairings of the air bubbles in the successive images.

7. The bubble measurement system of claim 1 , wherein the pairing module performs a binary image classification analysis when the number of air bubbles in successive images differs, the binary image classification analysis determining if a bubble overlap condition has occurred.

8. The bubble measurement system of claim 7 , wherein the pairing module includes a support vector machine (SVM) classifier to perform the binary image classification analysis.

9. The bubble measurement system of claim 7 , wherein the imaging controller includes a region-based convolutional neural network (R-CNN) to perform image analysis when the bubble overlap condition occurs to measure each of the overlapping air bubbles.

10. The bubble measurement system of claim 1 , wherein the imaging system includes a measurement database having inputs including an image ID for each image, a bubble ID for each bubble, and measurement data for each bubble in each image, the pairing module using the inputs from the measurement database to analyze the successive images.

11. The bubble measurement system of claim 1 , wherein the imaging system includes a validation module using the bubble measurements to validate operation of the bubble detector.

12. The bubble measurement system of claim 1 , wherein the imaging system includes a calibration module using the bubble measurements to calibrate operation of the bubble detector.

13. A bubble measurement system comprising:

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 having an imaging device for imaging the fluid and air bubbles in the flow path of the vessel of the bubble detector, the imaging system having an imaging controller coupled to the imaging device and receiving images from the imaging device, the imaging controller processing the images to measure bubble size of each air bubble passing through the bubble detector, the imaging controller including a pairing module comparing successive images and the air bubbles in successive images to measure all bubbles flowing through the vessel, the pairing module including a neural network algorithm to match air bubbles in the successive images;

wherein the pairing module performs a binary image classification analysis when the number of air bubbles in successive images differs, the binary image classification analysis determining if a bubble overlap condition has occurred; and

wherein the imaging controller includes a region-based convolutional neural network (R-CNN) to perform image analysis when the bubble overlap condition occurs to measure each of the overlapping air bubbles.

14. The bubble measurement system of claim 13 , wherein the pairing module compares successive images to determine when the air bubbles enter a field of view of the imaging device and when the air bubbles exit the field of view of the imaging device.

15. The bubble measurement system of claim 13 , wherein the pairing module uses positional data of the air bubbles in the successive images and uses measurement data of the air bubbles in the successive images to match the air bubbles between the successive images.

16. The bubble measurement system of claim 13 , wherein the neural network algorithm is a Gale-Shapley algorithm to find optimal pairings of the air bubbles in the successive images.

17. The bubble measurement system of claim 13 , wherein the pairing module includes a support vector machine (SVM) classifier to perform the binary image classification analysis.

18. The bubble measurement system of claim 13 , wherein the imaging controller includes a region-based convolutional neural network (R-CNN) to perform image analysis when the bubble overlap condition occurs to measure each of the overlapping air bubbles.

19. The bubble measurement system of claim 13 , wherein the imaging system includes a measurement database having inputs including an image ID for each image, a bubble ID for each bubble, and measurement data for each bubble in each image, the pairing module using the inputs from the measurement database to analyze the successive images.

20. A method of measuring air bubbles in fluid flowing through a vessel, the method comprising:

imaging the fluid and the air bubbles passing through a bubble detector using an imaging device;

processing the images of the air bubbles to determine bubble sizes of the air bubbles; and

comparing successive images and the air bubbles in successive images to measure total volume of all bubbles flowing through the vessel and to identify when bubble overlap conditions occur.

21. The method of claim 20 , further comprising using a neural network algorithm to match air bubbles in the successive images.

22. The method of claim 20 , further comprising performing a binary image classification analysis when the number of air bubbles in successive images differs to determine if a bubble overlap condition has occurred.

23. The method of claim 22 , further comprising performing image analysis using a region-based convolutional neural network (R-CNN) when the bubble overlap condition occurs to measure each of the overlapping air bubbles.

24. The method of claim 20 , validating operation of the bubble detector based on the measured total volume of all of the bubbles flowing through the vessel.

25. The method of claim 20 , calibrating operation of the bubble detector based on the measured total volume of all of the bubbles flowing through the vessel.

Assignments (2)
MERGER Recorded Jun 7, 2022
From: TE CONNECTIVITY SERVICES GMBH
To: TE CONNECTIVITY SOLUTIONS GMBH
Reel/Frame 060305/0923 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 4, 2022
From: WANG, SHU; WEN, DU; LU, ROBERTO FRANCISCO-YI; ZHOU, JIANKUN
To: TE CONNECTIVITY SERVICES GMBH
Reel/Frame 058540/0618 →