IP Library Granted Patent US 11,068,700
Granted Patent B2
US 11,068,700 · App. 16/431,374 · Granted Jul 20, 2021

Polarization imaging for facial recognition enhancement system and method

Inventors: J. Larry Pezzaniti (Huntsville, AL); David B. Chenault (Huntsville, AL)
Assignee: POLARIS SENSOR TECHNOLOGIES, INC.
G06K9/00288G02B5/201G02B27/00G02B27/288G06K9/00255G06K9/00275G06K9/209G06K9/4661
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 11,068,700
App. No.
16/431,374
Granted
Jul 20, 2021
Kind
B2
Abstract

A method for enhancing an image for facial recognition comprises capturing an image of the face with a polarizer and correcting the polarized image for non-uniformity. Stokes Parameters S 0 , S 1 , S 2 are obtained by weighted subtraction of the polarized image to form Stokes images. DoLP is computed from the Stokes images, and facial recognition algorithms are applied to the DoLP image. A system for enhancing images for facial recognition comprises a polarimeter configured to record polarized image data of a subject's face, a signal processing unit and logic configured to receive and store in memory the image data from the polarimeter, calculate Stokes parameters from the image data, and compute a DoLP image from the Stokes parameters.

Claims (27)

1. A method of obtaining enhanced images for facial recognition, the method comprising:

recording, with a long wave infrared polarimeter, raw image data of a subject's face to obtain a number of polarized images;

correcting the polarized images for non-uniformity;

calculating Stokes parameters S 0 , S 1 , and S 2 from the polarized images to create Stokes images by weighted subtraction of the polarized images; and

computing polarization images derived from the Stokes images.

2. The method of claim 1 , wherein the step of computing polarization images derived from the Stokes images comprise computing a Degree of Linear Polarization (DoLP) image from the Stokes images.

3. The method of claim 1 , further comprising applying facial recognition algorithms to polarization images.

4. The method of claim 2 , wherein the polarimeter comprises an objective lens, a polarization filter, and a focal plane array.

5. The method of claim 4 , wherein the focal plane array comprises a plurality of wire grid polarizers.

6. The method of claim 4 , wherein the focal plane array comprises a plurality of super pixels, each super pixel comprised of a 2×2 array that measures four (4) states of polarization, the states of polarization comprising 0°, 45°, 90°, and 135° states of linearly polarized light.

7. The method of claim 4 , wherein the focal plane array comprises a plurality of super pixels, each super pixel comprised of a three-pixel array that measures three (3) states of polarization, the states of polarization comprising 0°, 60° and 120° states of linearly polarized light.

8. The method of claim 4 , further comprising rotating the polarization filter and capturing images of the face with the filter at multiple orientations.

9. The method of claim 8 , wherein the multiple orientations comprise 0°, 45°, 90°, and 135°.

10. The method of claim 9 , wherein the images obtained at the multiple orientations are captured sequentially in time.

11. The method of claim 10 , wherein the images obtained at the multiple orientations are captured before the subject's face moves more than ¼ pixel.

12. A system for enhancing images for facial recognition, comprising:

a long wave infrared polarimeter configured to record polarized image data of a subject's face;

the system configured to record, with the long wave infrared polarimeter, raw image data of a subject's face to obtain a number of polarized images, the system further configured to correct the polarized images for non-uniformity, to calculate Stokes parameters S 0 , S 1 , and S 2 from the polarized images to create Stokes images by weighted subtraction of the polarized images; and to compute polarization images derived from the Stokes images.

13. The system of claim 12 , the system further configured to compute a Degree of Linear Polarization (DoLP) image from the Stokes parameters and apply facial recognition algorithms to the DoLP image.

14. The system of claim 12 , the polarimeter comprising an objective lens, a polarization filter, and a focal plane array.

15. The system of claim 14 , wherein the focal plane array comprises a plurality of wire grid polarizers.

16. The system of claim 14 , wherein the focal plane array comprises a plurality of super pixels, each super pixel comprised of a 2×2 array that measures four (4) states of polarization, the states of polarization comprising 0°, 45°, 90°, and 135° states of linearly polarized light.

17. The system of claim 14 , wherein the focal plane array comprises a plurality of super pixels, each super pixel comprised of a three-pixel array that measures three (3) states of polarization, the states of polarization comprising 0°, 60° and 120° states of linearly polarized light.

18. The system of claim 14 , wherein the polarization filter is a rotating filter and the logic is further configured to capture images of the face at multiple orientations of the rotating filter.

19. The system of claim 17 , wherein the multiple orientations comprise 0°, 45°, 90°, and 135°.

20. The system of claim 18 , further configured to capture the images of the multiple orientations sequentially in time.

21. The system of claim 19 , further configured to capture the images of the multiple orientations before the subject's face moves more than ¼ pixel.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jun 4, 2019
From: PEZZANITI, J. LARRY; CHENAULT, DAVID B.
To: POLARIS SENSOR TECHNOLOGIES, INC.
Reel/Frame 049365/0823 →
Continuity (3)
Continuation 14602823 · Jan 22, 2015
Provisional Application 61930272 · Jan 22, 2014
Related Publication 20200082159A1 · Mar 12, 2020