Biometrics based on locally consistent features
View Patent ↗Systems, devices, methods, and software are described for biometric sensors that permit a reduction in the size of the sensing area without significant reduction in biometric functionality of the sensor. A skin site of an individual is illuminated, and light scattered from the skin site is received. An image of a locally consistent feature of the skin site is formed from the received light. The locally consistent feature is analyzed to perform a biometric function.
1. A method of performing a biometric function, the method comprising:
illuminating a small-area purported skin site of an individual with illumination light;
receiving light scattered from the small-area purported skin site;
generating a local feature profile from the received light, wherein the local feature profile identifies a feature of the small-area purported skin site of a type predetermined to exhibit substantial local consistency across non-overlapping skin sites of the individual; and
analyzing the generated local feature profile to perform the biometric function, the biometric function selected from a group consisting of an identity function, a demographic function, and an anthropometric function,
wherein analyzing the generated local feature profile comprises comparing the generated local feature profile with a reference local feature profile generated from light scattered from a small-area reference skin site of the individual that was previously captured during biometric enrollment of the individual and does not overlap with the small-area purported skin site.
2. The method recited in claim 1 , wherein generating the local feature profile comprises:
forming an image from the received light;
generating spatially-distributed multispectral data from the image; and
processing the spatially-distributed multispectral data to generate the local feature profile.
3. The method recited in claim 1 , wherein generating the local feature profile comprises:
forming an image from the received light;
generating an image-texture measure from the image; and
processing the generated image-texture measure to generate the local feature profile.
4. The method recited in claim 1 wherein:
analyzing the generated local feature profile comprises determining an identity of the individual from the generated local feature profile.
5. The method recited in claim 1 wherein:
analyzing the generated local feature profile comprises estimating a demographic characteristic of the individual from the generated local feature profile.
6. The method recited in claim 1 wherein:
analyzing the generated local feature profile comprises estimating an anthropometric characteristic of the individual from the generated local feature profile.
7. The method recited in claim 1 wherein illuminating the small-area purported skin site comprises illuminating the small-area purported skin site under a plurality of distinct optical conditions.
8. The method recited in claim 7 wherein illuminating the small-area purported skin site under the plurality of distinct optical conditions comprises illuminating the small-area purported skin site with light under a plurality of distinct polarization conditions.
9. The method recited in claim 7 wherein illuminating the small-area purported skin site under the plurality of distinct optical conditions comprises illuminating the small-area purported skin site with light under a plurality of distinct wavelengths.
10. The method recited in claim 7 wherein generating the local feature profile from the received light comprises:
generating a plurality of images, each of the plurality of images corresponding to one of the distinct optical conditions; and
applying a compositing algorithm to the plurality of images.
11. A biometric sensor comprising:
an illumination subsystem disposed to illuminate a small-area purported skin site of an individual with illumination light;
a detection subsystem disposed to receive light scattered from the small-area purported skin site; and
a computational unit interfaced with the detection subsystem and having:
instructions for generating a local feature profile from the received light, wherein the local feature profile identifies a feature of the small-area purported skin site of a type predetermined to exhibit substantial local consistency across non-overlapping skin sites of the individual; and
instructions for analyzing the generated local feature profile to perform a biometric function selected from a group consisting of an identity function, a demographic function, and an anthropometric function,
wherein analyzing the generated local feature profile comprises comparing the generated local feature profile with a reference local feature profile generated from light scattered from a small-area reference skin site of the individual that was previously captured during biometric enrollment of the individual and does not overlap with the small-area purported skin site.
12. The biometric sensor recited in claim 11 , wherein the instructions for generating the local feature profile comprise:
forming an image from the received light;
generating spatially-distributed multispectral data from the image; and
processing the spatially-distributed multispectral data to generate the local feature profile.
13. The biometric sensor recited in claim 11 , wherein the instructions for generating the local feature profile comprise:
forming an image from the received light;
generating an image-texture measure from the image; and
processing the generated image-texture measure to generate the local feature profile.
14. The biometric sensor recited in claim 11 wherein:
the instructions for generating the local feature profile comprise determining an identity of the individual from the generated local feature profile.
15. The biometric sensor recited in claim 11 wherein:
the instructions for generating the local feature profile comprise estimating a demographic characteristic of the individual from the generated local feature profile.
16. The biometric sensor recited in claim 11 wherein:
the instructions for generating the local feature profile comprise estimating an anthropometric characteristic of the individual from the generated local feature profile.
17. The biometric sensor recited in claim 11 wherein the illumination subsystem is disposed to illuminate the small-area purported skin site under a plurality of distinct optical conditions.
18. The biometric sensor recited in claim 17 wherein the plurality of distinct optical conditions comprises a plurality of distinct polarization conditions.
19. The biometric sensor recited in claim 17 wherein the plurality of distinct optical conditions comprises light under a plurality of distinct wavelengths.
20. The biometric sensor recited in claim 17 wherein the instructions for generating the local feature profile from the received light comprise:
instructions for generating a plurality of images, each of the plurality of images corresponding to one of the distinct optical conditions; and
instructions for applying a compositing algorithm to the plurality of images.
21. The method recited in claim 1 , wherein the feature of the small-area purported skin site comprises artificial pigmentation.
22. The method recited in claim 21 , wherein analyzing the generated local feature profile comprises verifying a presence, quantity, and/or shape of the artificial pigmentation.
23. The biometric sensor recited in claim 11 , wherein the feature of the small-area purported skin site comprises artificial pigmentation.
24. The biometric sensor recited in claim 23 , wherein the instructions for analyzing the generated local feature profile comprise verifying a presence, quantity, and/or shape of the artificial pigmentation.