Method of performing shape localization

• US 20060008149A1
• Filed: 07/12/2004
• Published: 01/12/2006
• Est. Priority Date: 07/12/2004
• Status: Active Grant

• Current Assignee: Board of Trustees of The University of Illinois
• Original Assignee: Board of Trustees of The University of Illinois
• Inventors: Tu, Jilin, Huang, Thomas S.
• Granted Patent: US 7,454,039 B2
• Time in Patent Office: Days
• US Class Current: 382/190
• CPC Class Codes: G06V 10/7553 based on shape, e.g. active shape models [ASM]
• CPC Class Codes: G06V 40/165 using facial parts and geometric relationships

Abstract

A method for performing shape localization in an image includes deriving a model shape from a database of a plurality of sample shapes. The model shape is defined by a set of landmarks. The method further includes deriving a texture likelihood model of present sub-patches of the set of landmarks defining the model shape in the image, and proposing a new set of landmarks that approximates a true location of features of the shape based on a sample proposal model of the present sub-patches. A CONDENSATION algorithm is used to derive the texture likelihood model and the proposed new set of landmarks.

Claims

1. A method for performing shape localization in an image, comprising:

deriving a model shape from a database of a plurality of sample shapes, said model shape being defined by a set of landmarks;

deriving a texture likelihood model of present sub-patches of said set of landmarks defining said model shape in the image; and

proposing a new set of landmarks that approximates a true location of features of the shape based on a sample proposal model of said present sub-patches;

wherein said deriving said texture likelihood model and said proposing said new set of landmarks are conducted using a CONDENSATION algorithm.

2. The method as defined in claim 1, wherein

said model shape is derived from a prior probabilistic distribution of a predefined model p(m),

said texture likelihood model of said present sub-patches is derived from a local texture likelihood distribution model p(I|m), and

said sample proposal model is derived based on a texture likelihood model of subsequent sub-patches of a set of landmarks in the image at proposed locations in a vicinity of said present sub-patches of said present set of landmarks.

3. The method as defined in claim 2, wherein the a prior probabilistic distribution of a predefined model p(m) describes a 2-dimensional shape represented as a vector,

S=(x₁, x₂, . . . ,x_K,y₁,y₂, . . . ,y_K)^T

of length 2K, where K is a number of landmarks that define said shape.

4. The method as defined in claim 3, wherein said shape is modeled as a vector

5. The method as defined in claim 4, wherein the a prior probabilistic distribution of a predefined model p(m) is obtained by learning a mixture of Gaussian model after projecting said shape vector S in the k dimensional active shape model (ASM) eigenspace.

6. The method as defined in claim 4 wherein, said shape vector S is rearranged as

7. The method as defined in claim 6 wherein, said shape vector S is denoted as

8. The method as defined in claim 2, wherein the local texture likelihood distribution model p(I|m) is defined as

9. The method as defined in claim 8, wherein

I_i={Γ

₁⁽ⁱ⁾,Γ

₂⁽ⁱ⁾, . . .Γ

_K⁽ⁱ⁾}

is a collection of local observation of shape features in image at interation i, and

10. The method as defined in claim 9, wherein the formula

11. The method as defined in claim 10, wherein said model parameter space equation is obtained from new model sample is proposed as

12. The method as defined in claim 1, wherein the CONDENSATION algorithm is performed separately on at least one reduced-size of the image prior to performing the CONDENSATION algorithm on a full-size of the image.

13. The method as defined in claim 12, wherein the CONDENSATION algorithm is performed separately for a plurality of image resolutions starting from a low image resolution to a high image resolution.

14. The method as defined in claim 13, wherein the number of landmarks defining said model shape is increased from the low image resolution to the high image resolution.

15. The method as defined in claim 14, wherein dimension of shape eigen-space are also increased from the low image resolution to the high image resolution.

16. The method as defined in claim 12, wherein the CONDENSATION algorithm is performed hierarchically on a plurality of image resolutions starting from a low image resolution to a high image resolution.

17. Method for performing a face localization in an image based on a Bayesian rule, comprising:

deriving a predefined face shape model m;

employing conditional density propagation (CONDENSATION) algorithm to locate a face shape in the image using a prior probabilistic distribution of a model p(m) based on said predefined face shape model m, and a local texture likelihood distribution given said predefined face shape model with specific model parameters p(I|m).

18. The method as defined in claim 17, wherein the CONDENSATION algorithm is performed separately on at least one reduced-size of the image prior to performing the CONDENSATION algorithm on a full-size of the image.

19. The method as defined in claim 17, wherein the a prior probabilistic distribution of a predefined model p(m) describes a 2-dimensional face shape represented as a vector,

S=(x₁, x₂, . . . , x_K, y₁, y₂, . . . y_K)^T

of length 2K, where K is a number of landmarks that define said face shape.

20. The method as defined in claim 19, wherein the face shape is modeled as a vector

21. The method as defined in claim 20, wherein the a prior probabilistic distribution of a predefined model p(m) is obtained by learning a mixture of Gaussian model after projecting said face vector S in the k dimensional active shape model (ASM) eigenspace.

22. The method as defined in claim 21 wherein, said face shape vector S is rearranged as

23. The method as defined in claim 22 wherein, said shape vector S is denoted as

24. The method as defined in claim 17, wherein the local texture likelihood distribution model p(I|m) is defined as

25. The method as defined in claim 24, wherein

I_i32 {Γ

₁⁽ⁱ⁾,Γ

₂⁽ⁱ⁾, . . . , Γ

_K⁽ⁱ⁾}

is a collection of local observation of facial features in image at interation i, and

26. The method as defined in claim 25, wherein the formula

27. The method as defined in claim 26, wherein said model parameter space equation is obtained from new model sample is proposed as