Method and apparatus for estimating body shape

Patent metadata

Field	Value
Publication number	US-10339706-B2
Application number	US-201816008637-A
Country	US
Kind code	B2
Filing date	Jun 14, 2018
Priority date	Aug 15, 2008
Publication date	Jul 2, 2019
Grant date	Jul 2, 2019

Abstract

A system and method of estimating the body shape of an individual from input data such as images or range maps. The body may appear in one or more poses captured at different times and a consistent body shape is computed for all poses. The body may appear in minimal tight-fitting clothing or in normal clothing wherein the described method produces an estimate of the body shape under the clothing. Clothed or bare regions of the body are detected via image classification and the fitting method is adapted to treat each region differently. Body shapes are represented parametrically and are matched to other bodies based on shape similarity and other features. Standard measurements are extracted using parametric or non-parametric functions of body shape. The system components support many applications in body scanning, advertising, social networking, collaborative filtering and Internet clothing shopping.

First claim

What is claimed is:

1. A method comprising:
obtaining data representing a body of an individual in a first pose and a second pose, wherein the data comprises one of image data of the body captured via a camera and partial depth information of the body captured via a range sensor, and wherein the first pose differs from the second pose; and
estimating a consistent body shape of the individual across the first pose and the second pose by fitting a parametric body model of the body to the data to generate a set of pose parameters and a set of shape parameters, wherein the estimating separately factors (1) changes in a shape of the body due to changes between the first pose and the second pose to yield the set of pose parameters from (2) changes in the body of the individual due to identity to yield the set of shape parameters.

2. The method of claim 1 , wherein the fitting of the parametric body model to the data comprises executing an objective function defined at least in part by the set of pose parameters and the set of shape parameters.

3. The method of claim 2 , wherein the objective function processes first data from the first pose and second data from the second pose and implements the consistent body shape across the first pose and the second pose.

4. The method of claim 1 , wherein the individual has at least one of an associated gender and an associated ethnicity, and wherein the fitting of the parametric body model to the data comprises processing an objective function defined at least in part by the set of pose parameters, the set of shape parameters, and a specified parameter corresponding to at least one of the associated gender and the associated ethnicity of the individual.

5. The method of claim 3 , wherein the data represents an at least partially clothed body of the individual, wherein the estimating further comprising estimating a body shape of a portion of the at least partially clothed body that is covered by at least one piece of clothing of the at least partially clothed body based on the parametric body model and the data representing the at least partially clothed body.

6. The method of claim 5 , wherein estimating the body shape further comprises detecting, via image classifiers, regions corresponding to at least one of skin, hair, and clothing.

7. The method of claim 6 , wherein the fitting of the parametric body model of the at least partially clothed body utilizes an objective function that permits the estimation to be substantially within the second data.

8. The method of claim 1 , wherein the parametric body model is a statistical parametric body model.

9. A system comprising:
a processor; and
a computer-readable storage medium storing instructions which, when executed by the processor, cause the processor to perform operations comprising:
obtaining data representing a body of an individual in a first pose and a second pose, wherein the data comprises one of image data of the body captured via a camera and partial depth information of the body captured via a range sensor, and wherein the first pose differs from the second pose; and
estimating a consistent body shape of the individual across the first pose and the second pose by fitting a parametric body model of the body to the data to generate a set of pose parameters and a set of shape parameters, wherein the estimating separately factors (1) changes in a shape of the body due to changes between the first pose and the second pose to yield the set of pose parameters from (2) changes in the body of the individual due to identity to yield the set of shape parameters.

10. The system of claim 9 , wherein the fitting of the parametric body model to the data comprises processing an objective function defined at least in part by the set of pose parameters and the set of shape parameters.

11. The system of claim 10 , wherein the objective function processes first data from the first pose and second data from the second pose and implements the consistent body shape across the first pose and the second pose.

12. The system of claim 9 , wherein the individual has at least one of an associated gender and an associated ethnicity, and wherein the fitting of the parametric body model to the data comprises processing an objective function defined at least in part by the set of pose parameters, the set of shape parameters, and a specified parameter corresponding to at least one of the associated gender and the associated ethnicity of the individual.

13. The system of claim 11 , wherein the data represents an at least partially clothed body of the individual, wherein the estimating further comprising estimating a body shape of a portion of the at least partially clothed body that is covered by at least one piece of clothing of the at least partially clothed body based on the parametric body model and the data representing the partially clothed body.

14. The system of claim 13 , wherein estimating the body shape further comprises detecting, via image classifiers, regions corresponding to at least one of skin, hair, and clothing.

15. The system of claim 14 , wherein the fitting of the parametric body model of the at least partially clothed body utilizes an objective function that permits the estimation to be substantially within the second data.

16. The system of claim 9 , wherein the parametric body model is a statistical parametric body model.

17. A computer-readable storage device storing instructions which, when executed by a processor, cause the processor to perform operations comprising:
obtaining data representing a body of an individual in a first pose and a second pose, wherein the data comprises one of image data of the body captured via a camera and partial depth information of the body captured via a range sensor, and wherein the first pose differs from the second pose; and
estimating a consistent body shape of the individual across the first pose and the second pose by fitting a parametric body model of the body to the data to generate a set of pose parameters and a set of shape parameters, wherein the estimating separately factors (1) changes in a shape of the body due to changes between the first pose and the second pose to yield the set of pose parameters from (2) changes in the body of the individual due to identity to yield the set of shape parameters.

18. The computer-readable storage device of claim 17 , wherein the fitting of the parametric body model to the data comprises processing an objective function defined at least in part by the set of pose parameters and the set of shape parameters.

19. The computer-readable storage device of claim 18 , wherein the objective function processes first data from the first pose and second data from the second pose and implements the consistent body shape across the first pose and the second pose.

20. The computer-readable storage device of claim 18 , wherein the individual has at least one of an associated gender and an associated ethnicity, and wherein the fitting of the parametric body model to the data comprises processing an objective function defined at least in part by the set of pose parameters, the set of shape parameters, and a specified parameter corresponding to at least one of the associated gender and the associated ethnicity of the individual.

Assignees

• Univ Brown

Inventors

• Black Michael J
• Balan Alexandru O
• Weiss Alexander W
• Sigal Leonid
• Loper Matthew M
• St Clair Timothy S

Classifications

• G06V10/763

  Non-hierarchical techniques, e.g. based on statistics of modelling distributions · CPC title

• G06F18/2321

  using statistics or function optimisation, e.g. modelling of probability density functions · CPC title

• G06T7/77

  using statistical methods · CPC title

• G06Q30/0601

  Electronic shopping [e-shopping] · CPC title

• G06T7/75

  involving models · CPC title