Image processing apparatus and method

1. An image processing apparatus, comprising:

a first processor to generate a first texture by synthesizing a template texture and a facial area extracted from an input color image;

a second processor to generate a first mesh model by matching both a template mesh model corresponding to the template texture and a depth image corresponding to the input color image; and

a three dimensional (3D) model generator to generate a 3D facial model based on the first texture and the first mesh model.

2. The image processing apparatus of claim 1 , wherein the first processor comprises:

an extractor to extract a characteristic point of the input color image identifying the facial area within the input color image;

a first calculator to calculate a first transform function between the characteristic point of the input color image and a characteristic point of the template texture; and

a texture synthesizer to generate the first texture by synthesizing the facial area and the template texture through an application of the first transform function to pixels of the facial area of the input color image.

3. The image processing apparatus of claim 2 , wherein the extractor extracts the characteristic point of the input color image using an active shape model.

4. The image processing apparatus of claim 2 , wherein the first calculator calculates the first transform function using a radial basis function network.

5. The image processing apparatus of claim 2 , wherein the texture synthesizer adjusts a color tone of the template texture using a color tone of the facial area of the input color image.

6. The image processing apparatus of claim 1 , wherein the second processor comprises:

a second calculator to calculate a second transform function between a characteristic point of the template mesh model corresponding to the template texture and a characteristic point of the depth image corresponding to the input color image; and

a mesh matching unit to generate the first mesh model by matching the template mesh model and the depth image through an application of the second transform function to an entire set of vertexes of the template mesh model.

7. The image processing apparatus of claim 6 , wherein the second calculator calculates the second transform function using a radial basis function network.

8. The image processing apparatus of claim 1 , wherein the 3D model generator extracts a hair portion from the input color image, selects, from a plurality of pre-stored hair models, a hair model having a hair portion similar to the extracted hair portion, and synthesizes the selected hair model with the 3D facial model.

9. An image processing apparatus comprising:

a first processor to generate a first texture by synthesizing a template texture and a facial area extracted from an input color image;

a second processor to transform, to a first mesh model, a template mesh model corresponding to the template texture; and

a 3D model generator to generate a 3D facial model based on the first mesh model and the template texture.

10. An image processing method, comprising:

generating, using a processor, a first texture by synthesizing a template texture and a facial area extracted from an input color image;

generating, using a processor, a first mesh model by matching both a template mesh model corresponding to the template texture and a depth image corresponding to the input color image; and

generating a 3D facial model based on the first texture and the first mesh model.

11. The image processing method of claim 10 , wherein the generating of the first texture comprises:

extracting a characteristic point of the input color image identifying the facial area within the input color image;

calculating a first transform function between the characteristic point of the input color image and a characteristic point of the template texture; and

generating the first texture by synthesizing the facial area and the template texture through an application of the first transform function to pixels of the facial area of the input color image.

12. The image processing method of claim 11 , wherein the characteristic point of the input color image is extracted using an active shape model.

13. The image processing method of claim 11 , wherein the first transform function is calculated using a radial basis function network.

14. The image processing method of claim 11 , wherein the generating of the first texture further comprises:

adjusting a color tone of the template texture using a color tone of the facial area of the input color image.

15. The image processing method of claim 10 , wherein the generating of the first mesh model comprises:

calculating a second transform function between a characteristic point of the template mesh model corresponding to the template texture and a characteristic point of the depth image corresponding to the input color image; and

generating the first mesh model by matching the template mesh model and the depth image through an application of the second transform function to an entire set of vertexes of the template mesh model.

16. The image processing method of claim 15 , wherein the second transform function is calculated using a radial basis function network.

17. The image processing method of claim 10 , wherein the generating of the 3D facial model comprises:

extracting a hair portion from the input color image;

selecting, from a plurality of pre-stored hair models, a hair model having a hair portion similar to the extracted hair portion; and

synthesizing the selected hair model with the 3D facial model.

18. An image processing method, comprising:

Generating, using a processor, a first texture by synthesizing a template texture and a facial area extracted from an input color image; and

Generating, using a processor, a 3D facial model based on the first texture and a template mesh model corresponding to the template texture.

19. The image processing method of claim 18 , wherein the generating of the 3D facial model comprises:

transforming, to a first mesh model, the template mesh model corresponding to the template texture; and

generating the 3D facial model based on the first mesh model and the template texture.

20. A non-transitory computer readable recording medium storing a program to cause at least one processing element to implement the method of claim 10 .