Methods and systems for establishing accurate phenotype metrics

Patent metadata

Field	Value
Publication number	US-12333765-B2
Application number	US-202418639514-A
Country	US
Kind code	B2
Filing date	Apr 18, 2024
Priority date	Oct 27, 2022
Publication date	Jun 17, 2025
Grant date	Jun 17, 2025

Abstract

Systems and methods for measuring skin tone. A camera and computer may measure skin tone health of a face dataset and color quality of a single face image. The computer and camera may measure skin tone across wide sample of volunteers, express skin tone in CIELCH coordinates; and identify the range of observed lightness, chromaticity, and hue values in the population to determine a sample set. The computer and camera may be configured to match the image to a coordinate system of lightness, chromaticity, and hue in the sample set. If a match is not possible, the computer may be configured to select a coordinate closest to the measured lightness, chromaticity, and hue.

First claim

What is claimed is:

1. A process flow to create a data driven skin tone target comprising:
generating a face data set comprising a set of images of faces;
providing a camera comprising camera properties associated with capturing an image;
measuring skin tone across the face data set;
expressing skin tone of the images in CIELAB or CIELCH coordinates;
identifying a range of observed lightness, chromaticity, and hue values in the images of faces;
identifying a set of N colors that span the range of observed lightness, chromaticity, and hue values where N is a natural number;
spectrally printing N color patches to create a data-driven skin tone calibration target;
using the data-driven skin tone calibration target to adjust the camera properties to reliably reproduce human skin tones;
defining a range of HCL values;
if the HCL values are inside the range, then setting a color health value of the face equal to 1; else determining a shortest difference in CIELAB space to a boundary of natural color range and setting the color health value of the face proportional to the shortest difference in CIELAB space to the boundary of the natural color range; and
measuring a proportion of faces in the data set that have color health values equal to 1.

2. The process flow of claim 1 wherein the face data set comprises at least twenty-face images of faces; the images of faces coming from a random assortment of individuals having diverse race, gender, and age.

3. The process flow of claim 1 comprising:
for each image in the face data set, quantifying skin tone; and
detecting face box and segment skin regions.

4. The process flow of claim 1 comprising:
determining skin pixels with specular reflection in the face data set; and
determining skin pixels with diffuse reflection in the face data set.

5. The process flow of claim 4 comprising:
removing skin pixels with specular reflection;
averaging skin pixels with diffuse reflection; and
computing CIELCH coordinates for the faces.

6. A system configured to create a data driven skin tone target comprising:
a skin tone measuring device configured to measure skin tone across a sample population;
a camera comprising camera properties associated with capturing an image;
a computer connected to the skin tone measuring device;
the computer configured to identify a range of observed lightness, chromaticity, and hue values in the sample population;
the computer configured to:
express the skin tone in CIELAB or CIELCH coordinates;
identify a set of N colors spanning the range of observed lightness, chromaticity, and hue values, where N is a natural number;
spectrally print N color patches to create a data-driven skin tone calibration target;
use the calibration target to adjust camera properties to reproduce human skin tones;
define a range of HCL values;
if the HCL values are inside the range, then set a color health value of the face equal to 1; else determine a shortest difference in CIELAB space to a boundary of natural color range and set the color health value of the face proportional to the shortest difference in CIELAB space to the boundary of the natural color range;
measure a proportion of faces in the data set that have color health values equal to 1; and
generate a face data set comprising a set of images of faces.

7. The system of claim 6 wherein the face data set comprises M images of faces, where M is a natural number.

8. The system of claim 6 wherein the faces are of individuals in a random distribution of race, gender, nationality, sex, and age.

9. The system of claim 6 wherein the computer is configured to select a subset of the face images in the data set using a randomization algorithm with a randomizer.

10. The system of claim 6 wherein the computer comprises a processor, memory, storage media, a motherboard, power supply, a network interface, non-transitory, computer readable code, software, and instructions for causing the processor to execute a series of processes.

11. The system of claim 6 wherein the computer is configured to define the range of values for HCL that fall within a standard deviation.

12. The system of claim 6 wherein the camera is configured to capture images of a plurality of users; the image may comprise an identifier.

13. The system of claim 12 wherein the computer is configured to determine data for a plurality of skin color calibration panels, the skin color calibration panels color composed of the measured lightness, chromaticity, and hue associated with an identifier.

14. The system of claim 13 wherein the computer is configured to determine data for a plurality of non-skin color calibration panels;
the non-skin color calibration panels color composed of the measured lightness, chromaticity, and hue associated with that identifier;
the skin color calibration panels comprising a lightness, chromaticity, and hue consistent with a lightness, chromaticity, and hue of human skin; and
the nonskin color calibration panels not comprising a lightness, chromaticity, and hue consistent with a lightness, chromaticity, and hue of human skin.

15. The system of claim 14 wherein the computer is configured to organize the skin color calibration panels and non-skin color calibration panels into a grid in memory of the computer.

16. The system of claim 15 wherein the computer is configured to export the grid into a tangible format or digital format; the exported grid being a colorchecker; and the colorchecker printed on a tangible surface comprising paper or plastic.

17. The system of claim 16 wherein the computer is configured to generate a color measuring tool configured to measure lightness, chromaticity, and hue values of the panels in the printed colorchecker or digital colorchecker so that the lightness, chromaticity, and hue values are known to the computer.

Assignees

• The Government Of The Us Secretary Of Homeland Security
• The Government Of The United States Of America Represented By The Secretary Of Homeland Security

Inventors

• Vemury Arun
• Sirotin Yevgeniy
• Howard John
• Tipton Jerry

Classifications

• G06T7/80

  Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration · CPC title

• G06T2207/10024

  Color image · CPC title

• G06T2207/30201

  Face · CPC title

• G06V40/171

  Local features and components; Facial parts (eye characteristics G06V40/18); Occluding parts, e.g. glasses; Geometrical relationships · CPC title

• H04N9/00

  Details of colour television systems · CPC title