Image recognition and analysis system and software

• US 8,254,657 B2
• Filed: 01/07/2008
• Issued: 08/28/2012
• Est. Priority Date: 01/08/2007
• Status: Expired due to Fees

• Current Assignee: Pollack Laboratories, Inc.
• Original Assignee: Pollack Laboratories, Inc.
• Inventors: Pollack, Michael J., Darnell, Branson J., Mandrachia, Steven J., Reichl, Gary, DiDomizio, Richard A.
• Primary Examiner(s): Vanore, David A
• Assistant Examiner(s): SMYTH, ANDREW P
• Application Number:
  US11/970,313
• Publication Number:
  US 20080166037A1
• Time in Patent Office: 1,695 Days
• Field of Search: None
• US Class Current: 382/133
• CPC Class Codes: G01N 15/0227 using imaging; using holography
• CPC Class Codes: G01N 15/1433 using image recognition
• CPC Class Codes: G01N 2015/0092 Monitoring flocculation or agglomeration
• CPC Class Codes: G01N 2015/025 Methods for single or grouped particles
• CPC Class Codes: G01N 2015/1497 Particle shape
• CPC Class Codes: G01N 21/8507 Probe photometers, i.e. with optical measuring part dipped into fluid sample
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/30024 Cell structures in vitro; Tissue sections in vitro
• CPC Class Codes: G06T 2207/30108 Industrial image inspection
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection

Abstract

A method of analyzing bubbles, cells, cell viability, or other particles or agglomerates in a process liquid contained in a vessel is provided. Images of bubbles, cells or other particles in the liquid are obtained in-situ with a vision probe extending through a wall of the vessel. The images are analyzed with image recognition software. The software measures at lease one of bubble, cell or particle size, mean diameter, surface area, flow rate, flow pattern, population distribution, viability, agglomerates or clumping, color change, viscosity, Sauter mean, ratio of surface area of bubbles relative to volume of bubbles, gas hold-up ratio of gas volume to volume of liquid, or interfacial area. The software distinguishes valid or viable bubbles, cells or particles that should be included in an analysis from invalid or non-viable bubbles, cells or particles that should not be included. The software can be configured to provide an analysis of the valid bubbles, cells or particles that fall within pre-set size and shape or viability parameters.

Claims

1. A method of analyzing bubbles, cells or other particles in a process liquid contained in a vessel, comprising the steps of:

obtaining images of the bubbles, cells or other particles in the liquid in-situ with a vision probe extending into the vessel;

generating a binary image for each image, each binary image having a background with spots located thereon corresponding to the bubbles, cells or particles appearing in the images;

analyzing the binary images with image recognition software to determine information concerning a process occurring within the vessel; and

adjusting a process condition of the process occurring within the vessel based on the information determined from said analyzing step;

said analyzing step including measuring at least one of bubble, cell or particle size, mean diameter, surface area, flow rate, flow pattern, population distribution, viability, agglomerates or clumping, color change, viscosity, Sauter mean, ratio of surface area of bubbles relative to volume of bubbles, gas hold-up ratio of gas volume to volume of liquid, and interfacial area;

said analyzing step including identifying valid or viable bubbles, cells or particles from invalid or non-viable bubbles, cells, or other particles appearing in the images by distinguishing the invalid bubbles, cells or particles that appear in the images and that are stuck to the vision probe relative to the valid bubbles, cells or particles that are freely flowing in the process liquid; and

said distinguishing step being accomplished by determining which bubbles, cells or particles appear in the same location in each of the images.

2. A method according to claim 1, wherein said images include at least about 100 frames of images.

3. A method according to claim 1, wherein the spots that appear in the binary images that are of a size less than a pre-determined number of pixels are deleted from the binary images.

4. A method according to claim 3, wherein binary images having annular or unfilled spots are filled in such that the spots in the binary images are solid.

5. A method according to claim 3, wherein any spots in the binary images contacting a border of the images are identified as invalid bubbles, cells or particles.

6. A method according to claim 5, wherein data including the location, size and shape of each spot corresponding to one of the valid bubbles, cells or particles is generated.

7. A method according to claim 6, wherein a report presenting the data is generated relative to the data corresponding to spots of a pre-determined shape and size.

8. A method of analyzing bubbles in a process liquid contained in a vessel, comprising the steps of:

releasing gas in the form of bubbles from a sparger in the process liquid within the vessel;

agitating the process liquid with an agitator within the vessel;

obtaining images of the bubbles in the process liquid in-situ with a vision probe extending into the vessel, the images being obtained by projecting light from the vision probe into the vessel to illuminate the gas bubbles and recording the images of illuminated bubbles with a camera mounted on the vision probe, the step of projecting light including emitting or reflecting light directionally from behind the gas bubbles relative to the camera to produce back lit images;

generating a binary image for each of the images, each binary image having a background with spots located thereon corresponding to the bubbles appearing in the image; and

analyzing the binary images with image recognition software;

said analyzing step including identifying valid or viable bubbles from invalid or non-viable bubbles appearing in the images;

said identifying of valid or viable bubbles including distinguishing the invalid bubbles that appear in the images and that are stuck to the vision probe relative to the valid bubbles that are freely flowing in the process liquid, the invalid bubbles being distinguished from the valid bubbles by determining which bubbles appear in the same location in each of the binary images; and

said analyzing step including measuring bubble size, mean diameter, surface area, flow rate, flow pattern, population distribution, viscosity, Sauter mean, ratio of surface area of bubbles relative to volume of bubbles, gas hold-up ratio of gas volume to volume of liquid, and interfacial area.

9. A method according to claim 8, wherein said analyzing step is performed in substantially real time, and further comprising the step of controlling a process condition in the vessel based on said analyzing step.

10. A method according to claim 9, wherein said step of controlling includes adjusting a flow rate of gas through the sparger, adjusting a rate of rotation of the agitator, adjusting temperature, or adjusting turbidity.

11. A method according to claim 8, wherein said step of obtaining images only occurs at preset intervals of time, and wherein light is only projected from the vision probe during the pre-set intervals.

12. A method according to claim 8, wherein the vision probe is flexible, and further comprising the step of re-positioning the vision probe within the vessel during an on-going process to observe a different region of the process fluid within the vessel, wherein said step of re-positioning the vision probe includes articulating the probe via a ball joint.