Spatial cross-correlating velocimeter

• US 5,249,238 A
• Filed: 03/19/1991
• Issued: 09/28/1993
• Est. Priority Date: 03/19/1991
• Status: Expired due to Fees

• Current Assignee: Narayanan M. Komerath, Philip A. Fawcett
• Original Assignee: Narayanan M. Komerath, Philip A. Fawcett
• Inventors: Komerath, Narayanan M., Fawcett, Philip A.
• Primary Examiner(s): Razavi, Michael T.
• Assistant Examiner(s): COUSO, YON JUNG
• Application Number:
  US07/671,529
• Time in Patent Office: 924 Days
• Field of Search: 382/42, 382/23, 382/17, 382/1, 73/861.06
• US Class Current: 382/107
• CPC Class Codes: G01P 5/001 Full-field flow measurement, e.g. determining flow velocity and direction in a whole region at the same time, flow visualisation
• CPC Class Codes: G01P 5/18 by measuring the time taken to traverse a fixed distance
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 7/223 using block-matching

Abstract

This invention enables the simultaneous measurement of two components of instantaneous velocity at several points over an area in a specified cross-section of a flowing fluid or on a vibrating solid surface. It is applicable to large areas, and to situations where the measurement must be conducted from large distances. It uses digital cross-correlation, between two images taken within a short time interval, of the scattering patterns from the plane of interest. The patterns can be random, periodic, or quasi-periodic, and the individual light-scattering features of the patterns need not be resolved. The information density in the images can vary from being sparse to dense. Since high values of correlation coefficients are not required, the method survives substantial amounts of noise in the images. The method is suitable for implementation with very short facility run-times. The post-processing is fully automated and is performed using moderate computational resources. The invention consists of the concept, the method, and the present implementations of the method.

Claims

1. A method of measuring velocities at a plurality of locations in a moving medium exhibiting imagable inhomogeneities that move with the medium, said method comprising the steps of:

(a) obtaining a first image of the inhomogeneities within a predetermined area of the medium;

(b) obtaining a second image of the inhomogeneities within the predetermined area of the medium with the second image being obtained within a time period sufficiently short to insure that substantially all of the inhomogeneities captured within the first image are present in the second image;

(c) digitizing the information contained in the first and second images obtained in steps (a) and (b);

(d) constructing a first numerical array representative of the digitized information produced in step (c) for the first image;

(e) constructing a second numerical array representative of the digitized information produced in step (c) for the second image;

(f) partitioning each of the first and second numerical arrays into a plurality of sub-arrays with each sub-array corresponding to one of the plurality of locations in the moving medium for which the velocity is to be measured;

(g) determining the cross-correlation matrix between each sub-array of the first numerical array and the corresponding sub-array of the second numerical array;

(h) identifying the extreme value within each of the cross-correlation matrices determined in step (g);

(i) for each cross-correlation matrix from step (g), determining the vector corresponding to the displacement within the matrix of the extreme value from the position corresponding to no displacement;

(j) converting each vector determined in step (i) to a vector having units of spacial displacement; and

(k) applying the known time between the first image and the second image to convert each displacement vector to a corresponding velocity vector.

2. The method of claim 1 and wherein steps (a) and (b) comprise projecting a sheet of light through the moving medium to illuminate the inhomogeneities within a substantially planar section of the medium and recording an image of a predetermined area of the illuminated planar section.

3. The method of claim 2 and wherein the step of recording an image of a predetermined area of the illuminated planar section comprises capturing the image on video tape.

4. The method of claim 2 and wherein the step of recording an image of a predetermined area of the illuminated planar section comprises capturing the image on film.

5. The method of claim 1 and wherein the moving medium comprises a moving fluid.

6. The method of claim 5 and wherein the inhomogeneities exhibited by the moving fluid comprise particulate matter within the fluid with the particulate matter forming patterns that move with the fluid.

7. The method of claim 5 and wherein the inhomogeneities exhibited by the moving fluid comprise density gradients within the fluid that move with the fluid.

8. The method of claim 1 and wherein the moving medium is a solid surface and wherein the inhomogeneities comprise imagable features on the surface.

9. The method of claim 1 and wherein the inhomogeneities within the medium are formed by a multitude of particles in the medium with the individual particles themselves being too small to be imaged in steps (a) and (b).

10. A method of determining selected dynamics within a preselected substantially planar region of a fluid flow comprising the steps of:

(a) inducing within the flow patterns that emulate the movement of fluid within the flow;

(b) obtaining a first image of the patterns within the preselected substantially planar region of the flow;

(c) delaying for a time sufficient to allow the patterns within the preselected substantially planar region to shift with the moving fluid by a predetermined amount;

(d) obtaining a second image of the patterns within the preselected substantially planar region of the flow;

(e) digitizing the first image to generate a first numerical matrix with the elements of the first matrix being indicative of a predetermined characteristic of the imaged patterns;

(f) digitizing the second image to generate a second numerical matrix with the elements of the second matrix being indicative of the predetermined characteristic of the imaged patterns;

(g) calculating the cross-correlation between the second numerical matrix and the first numerical matrix;

(h) identifying the location of the extreme value of the cross-correlation; and

(i) mathematically converting the identified location of the extreme value to units of the selected dynamics to be determined.

11. The method of claim 10 and wherein steps (e) and (f) further comprise subdividing the first and second numerical matrices into sub-matrices and performing the remaining steps of the method on each sub-matrix to determine the selected dynamics for each sub-matrix and thus for a plurality of locations within the preselected substantially planar region of the fluid flow.

12. The method of claim 10 and wherein at least one of the selected dynamics is displacement and wherein step (i) comprises determining the shift in location of the extreme value from a predetermined reference location in the cross-correlation and converting the shift to units of spacial displacement from the first image to the second image.

13. The method of claim 11 and wherein at least one of the selected dynamics is velocity and wherein step (i) further comprises applying the known delay time between the first and second images to convert the displacement into units of velocity.

14. The method of claim 10 and wherein step (a) comprises seeding the flow with a multitude of particles that are carried by the flow and that together form the patterns that emulate the movement of fluid within the flow.

15. The method of claim 10 and wherein steps (g) through (i) are performed by an appropriately programmed computer.

16. The method of claim 10 and wherein steps (e) and (f) the predetermined characteristic is the intensity of the image at the locations corresponding to each element of the image.

18. The apparatus of claim 16 and wherein said means for obtaining first and second images of the patterns within the preselected substantially planar region of the flow comprises a video camera.

19. The apparatus of claim 16 and wherein said means for digitizing comprises an electronic digitizing board.

17. An apparatus for determining selected dynamics within a preselected substantially planar region of a fluid flow comprising the steps of:

means for inducing within the flow patterns that emulate the movement of fluid within the flow;

means for obtaining a first image of the patterns within the preselected substantially planar region of the flow;

means for inducing a delay for a time sufficient to allow the patterns within the preselected substantially region to shift with the moving fluid by a predetermined amount;

means for obtaining a second image of the patterns within the preselected substantially planar region of the flow;

means for digitizing the first image to generate a first numerical matrix with the elements of the first matrix being indicative of a predetermined characteristic of the imaged patterns;

means for digitizing the second image to generate a second numerical matrix with the elements of the second matrix being indicative of the predetermined characteristic of the imaged patterns;

means for calculating the cross-correlation between the second numerical matrix and the first numerical matrix;

means for identifying the location of the extreme value of the cross-correlation; and

means for mathematically converting the identified location of the extreme value to units of the selected dynamics to be determined.