Vision-based operating method and system

• US 7,945,076 B2
• Filed: 05/02/2008
• Issued: 05/17/2011
• Est. Priority Date: 04/05/2002
• Status: Expired due to Fees

• Current Assignee: Morphotrust USA Incorporated (Advent International Corporation)
• Original Assignee: Identix, Inc. (Advent International Corporation)
• Inventors: DeLean, Bruno
• Primary Examiner(s): Hung; Yubin
• Application Number:
  US12/114,685
• Publication Number:
  US 20090097713A1
• Time in Patent Office: 1,110 Days
• Field of Search: 382115-118, 382/124, 382/209, 382/305, 707/6
• US Class Current: 382/118
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06V 10/757 Matching configurations of points or features
• CPC Class Codes: G06V 40/172 Classification, e.g. identification

Abstract

A vision-based operating system is provided in which a computer-based system accepts visual inputs and takes appropriate actions in a wide variety of environments. One environment is a security environment in which the vision-based operating system acquires an image of a face and determines, with an arbitrarily small probability of a false positive, whether the face is of a person who is authorized to enter the environment. The method employed may include acquiring a digital image for comparison to a reference image and identifying two or more groups of pixels in the acquired image. The pixel groups in the reference image that are the best match for the two or more pixels groups in the acquired image are selected. The relative locations of the selected pixel groups in the reference image and the two or more pixel groups in the acquired image are compared, and the probability that the relative locations occurred randomly is determined.

Claims

1. A method of determining whether an acquired image matches a reference image, comprising:

acquiring a digital image for comparison to a reference image;

identifying two or more groups of pixels in the acquired image;

selecting pixel groups in the reference image that are the best match for the two or more pixel groups in the acquired image;

comparing the relative locations of the selected pixel groups in the reference image and the two or more pixel groups in the acquired image; and

determining the probability that the relative locations occurred randomly.

2. The method of claim 1, further comprising:

comparing the probability of a random match to a threshold probability for concluding that the images match.

3. The method of claim 2, wherein the threshold probability is set to reduce the probability of a false conclusion to an arbitrarily low number.

4. The method of claim 1, wherein finding the closest pixel group uses a least squares calculation of differences between the pixel groups in the reference image and the two or more pixel groups in the acquired image.

5. The method of claim 1, further comprising:

upon acquiring the reference image and the acquired image, filtering the images.

6. The method of claim 5, wherein the filtering comprises:

determining the luminance of a pixel based on an average of values of neighboring pixels;

comparing the value of the luminance of the pixel to a threshold value;

setting the filtered value of the pixel to zero if the luminance is below the threshold; and

setting the filtered value of the pixel to one if the luminance is equal to or greater than the threshold value.

7. The method of claim 6, further comprising:

normalizing the acquired image and the reference image by locating known features in a predetermined location.

8. The method of claim 1, wherein acquiring the image is accomplished by a vision processing system having a vision-based operating system.

9. The method of claim 1, further comprising:

comparing the acquired image to more than one reference image.

10. The method of claim 1, wherein comparing the relative locations of the two or more pixel groups of the acquired image and the pixel groups of the reference image comprises:

defining a first vector between a first pixel group of the acquired image and a second pixel group of the acquired image;

defining a second vector between a first pixel group of the reference image and a second pixel group of the reference image; and

calculating an absolute value of the difference between the first vector and the second vector.

11. The method of claim 10, further comprising:

repeating the calculation for additional pixel groups.

12. The method of claim 1, further comprising:

comparing the probability of a random match to a threshold probability for concluding that the images match;

repeating the steps for a different reference image; and

allowing unlimited number of mismatches before allowing a match to one of the reference images.

13. The method of claim 1, further comprising:

upon determining a match, adding the acquired image to a database of reference images to facilitate future matches.

15. The method of claim 1, wherein the acquired digital image corresponds to a patch of skin.

14. A system for determining whether an acquired image matches a reference image, comprising:

a camera configured to acquire a digital image for comparison to a reference image;

an identification module configured to identify two or more groups of pixels in the acquired image;

a selection module configured to select pixel groups in the reference image that are the best match for the two or more pixel groups in the acquired image;

a comparison module configured to compare the relative locations of the selected pixel groups in the reference image to the two or more pixel groups in the acquired image; and

a calculator configured to determine the probability that the relative locations occurred randomly.

16. The system of claim 14, wherein the acquired digital image corresponds to a patch of skin.

17. The system of claim 16, further comprising:

a match module configured to compare the probability of a random match to a threshold probability for concluding that the images match.

18. The system of claim 17, wherein the threshold probability is set to reduce the probability of a false conclusion to an arbitrarily low number.

19. The system of claim 16, wherein finding the closest pixel group uses a least squares calculation of differences between the pixel groups in the reference image and the two or more pixel groups in the acquired image.

20. The system of claim 16, further comprising:

a filter configured to filter images.

21. The system of claim 20, further comprising:

a normalization module configured to normalize the acquired image and the reference image by locating known features in a predetermined location.

22. The system of claim 16, wherein acquiring the image is accomplished by a vision processing system having a vision-based operating system.

23. The system of claim 16, further comprising:

a comparison module configured to compare the acquired image to more than one reference image.

24. The system of claim 16, wherein:

the calculator is configured to compare the probability of a random match to a threshold probability for concluding that the images match; and

further including;

a reference image retrieval module configured to repeat the steps for a different reference image, wherein the system allows an unlimited number of mismatches before allowing a match to one of the reference images.

25. The system of claim 16, further comprising:

an image storage module configured to upon determining a match, adding the acquired image to a database of reference images to facilitate future matches.

26. A method of analyzing an image comprising:

acquiring a first surface area of a surface location of an image of a person to be analyzed;

retrieving a second surface area of the same surface location of a reference image;

identifying two or more groups of pixels in said first surface area;

selecting groups of pixels from said second surface area that are a best match for said two or more groups of pixels in said first surface area;

comparing the relative locations of said pixel groups in said first and second surface areas;

determining a probability that the relative locations occurred randomly; and

calculating a score representing a likelihood that said first and second surface areas are from the same person.