Method and apparatus for converting skin color of image

• US 8,452,091 B2
• Filed: 07/12/2006
• Issued: 05/28/2013
• Est. Priority Date: 08/08/2005
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Oh, Hyun-hwa, Kwak, Young-shin, Park, Du-sik, Kim, Chang-yeong
• Primary Examiner(s): DRENNAN, BARRY T
• Application Number:
  US11/484,718
• Publication Number:
  US 20070031033A1
• Time in Patent Office: 2,512 Days
• Field of Search: 382/118, 382/167, 382/274
• US Class Current: 382/167
• CPC Class Codes: G06T 2207/20132 Image cropping
• CPC Class Codes: G06T 2207/30201 Face
• CPC Class Codes: G06T 7/143 involving probabilistic approaches, e.g. Markov random field [MRF] modelling
• CPC Class Codes: G06T 7/90 Determination of colour characteristics
• CPC Class Codes: G06V 40/161 Detection; Localisation; Normalisation
• CPC Class Codes: H04N 1/628 Memory colours, e.g. skin or sky
• CPC Class Codes: H04N 9/643 Hue control means, e.g. flesh tone control

Abstract

An image-processing method that converts a skin color of an image is provided, the method including detecting a face region of an image, when a face region is detected in the image, verifying whether the face region is a face by analyzing the characteristics of the face region, and when the region is determined to be a face, extracting a skin region in the image, and converting the skin color of the extracted skin region into a desired skin color.

Claims

1. An image-processing method that converts a skin color of an image implemented by an image processing apparatus comprising a processor having computing device-executable instructions, the method comprising:

detecting a face region of an image using multi-view;

verifying that the face region contains a face by analyzing characteristics of the face region when a face region is detected;

establishing a skin color model when the face region is verified to contain the face, wherein the skin color model is created in a two-dimensional oval by using skin color that is created based on a pixel value of the detected face region and by displaying skin color of the face region in a color area;

extracting a skin region from the image using the skin color model, wherein when color of a pixel is classified as a skin color in the skin color model, the pixel is determined to be a pixel of the skin region; and

converting the skin color of the extracted skin region into a desired skin color,

wherein the multi-view is moving sub window in a screen, or using a template with diversely rotated faces, and

wherein when the image includes the detected face region, the image is a whole image including pixels inside and outside of the face region.

2. The method of claim 1, wherein the characteristics of the face region comprise a proportion of pixels corresponding to a skin color.

3. The method of claim 1, wherein the characteristics of the face region comprise a result obtained by comparing a symmetry of a face of the face region.

4. The method of claim 1, wherein the skin color model is further created by selectively using skin color that has been created based on standard skin color or skin color that is created based on a pixel value of the detected face region, and wherein the extracting includes extracting a pixel having a skin color using a standard skin color model when the skin color that has been created was created using the skin color based on the standard skin color.

5. The method of claim 1, wherein the converting comprises:

locating an extracted skin color from the extracted skin region in a color region;

calculating a target point for converting of the extracted skin color distributed in the color region by obtaining datum points of a standard color region and the color region; and

converting the skin color based on the calculated target point.

6. The method of claim 1, further comprising maintaining the image when a face region is not detected.

7. The method of claim 1, further comprising maintaining the image when the face region does not contain a face.

8. An image processing apparatus that converts a skin color of an image, comprising:

a processor;

a face determining unit, controlled by the processor, that determines an existence of a face by extracting a face region of an image, maintains the image when a face does not exist in the image, and outputs the face region when a face exists in the image; and

an image-converting unit, controlled by the processor, that receives the face region from the face-determining unit, establishes a skin color model when the face region is verified to contain the face, extracts a skin region from the image using the skin color model, wherein when color of a pixel is classified as a skin color in the skin color model, the pixel is determined to be a pixel of the skin region, and converts a skin color of the extracted skin region in the image into a desired skin color,

wherein when the image includes the face region, the image is a whole image including pixels inside and outside of the face region,

wherein the skin color model is created in a two-dimensional oval by using skin color that is created based on pixel value of the detected face region and by displaying skin color of the face region in a color area, and

wherein the face determining unit detects the face region of the image using multi-view, the multi-view is moving sub window in a screen, or using a template with diversely rotated faces.

9. The apparatus of claim 8, wherein the face-determining unit comprises:

a face-detecting unit that detects the face region of the image; and

a face-verifying unit which receives the face region, analyzes characteristics of the face region to verify that the face region contains a face, maintains the image when the face region does not contain a face, and outputs the face region to the image-converting unit when the face region contains a face, and

wherein the characteristics of the face region comprise a proportion of pixels corresponding to a skin color.

10. The apparatus of claim 8, wherein the face-determining unit comprises:

a face-detecting unit that detects the face region of the image; and

a face-verifying unit which receives the face region, analyzes characteristics of the face region to verify that the face region contains a face, maintains the image when the face region does not contain a face, and outputs the face region to the image-converting unit when the face region contains a face, and

wherein the characteristics of the face region comprise a result obtained by comparing a symmetry of a face.

11. The apparatus of claim 8, wherein the image-converting unit comprises:

a skin-region-detecting unit that receives the face region from the face-determining unit, and that extracts a skin region from the image based on the face region; and

a skin-color-converting unit that converts a skin color of the skin region in the image.

12. The apparatus of claim 11, wherein the skin-region-detecting unit generates the skin color model based on the face region and extracts pixels having the skin color using the skin color model.

13. The apparatus of claim 11, wherein the skin-color-converting unit distributes the skin color extracted by the skin region extracting unit in a color region, calculates a target point for converting the extracted skin color distributed in the color region by obtaining the datum points of a standard color region and the color region, and converts the skin color based on the detected target point.

14. An image-processing method implemented by an image processing apparatus comprising a processor having computing device-executable instructions, the method comprising:

detecting whether at least one face region is present in a received image, the received image being obtained using multi-view;

verifying that a face is present in the detected at least one face region;

establishing a skin color model when a face is present, wherein the skin color model is created in a two-dimensional oval by using skin color that is created based on a pixel value of the at least one detected face region and by displaying skin color of the face region in a color area;

extracting a skin pixel including a skin color from the received image using the skin color model, wherein when color of a pixel is classified as a skin color in the skin color model, the pixel is determined to be a pixel of the skin region; and

converting the skin color of the extracted skin pixel into a desired skin color based on the skin color model,

wherein the multiview is moving sub window in a screen, or using a template with diversely rotated faces, and

wherein when the received image includes at least one detected face region, the received image is a whole image including pixels inside and outside of the at least one face region.

15. The method of claim 14, further comprising, when at least one face region is not present in the received image, outputting the image without executing the extracting and converting.

16. The method of claim 14, further comprising, when a face is not present in the at least one face region, outputting the image without executing the extracting and converting.

17. The method of claim 14, wherein, in the verifying, the presence of a face in the at least one face region is verified based on a size of the detected at least one face region, a proportion of pixels corresponding to a skin color in the at least one face region, or geometrical characteristics of the at least one face region.

18. The method of claim 14, wherein the skin color model is a 2-dimension (2-D) oval skin color model or a 3 -D skin color model.

19. The method of claim 14, wherein the detecting uses an AdaBoost method, a neural network based method, or a statistical method.

20. A non-transitory computer-readable storage medium encoded with processing instructions for causing a processor to execute an image-processing method that converts a skin color of an image, the method comprising:

detecting a face region of an image using multi-view;

verifying that the face region contains a face by analyzing characteristics of the face region when a face region is detected;

establishing a skin color model when the face region is verified to contain the face, wherein the skin color model is created in a two-dimensional oval by using skin color that is created based on a pixel value of the detected face region and by displaying skin color of the face region in a color area;

extracting a skin region from the image using the skin color model, wherein when color of a pixel is classified as a skin color in the skin color model, the pixel is determined to be a pixel of the skin region; and

converting the skin color of the extracted skin region into a desired skin color based on the skin color model,

wherein the multi-view is moving sub window in a screen, or using a template with diversely rotating faces, and

wherein when the image includes the detected face region, the image is a whole image including pixels inside and outside of the face region.

21. A non-transitory computer-readable storage medium encoded with processing instructions for causing a processor to execute an image-processing method that converts a skin color of an image, the method comprising:

detecting whether at least one face region is present in a received image, the received image being obtained using multi-view;

verifying that a face is present in the detected at least one face region is a face;

establishing a skin color model when a face is present, wherein the skin color model is created in a two-dimensional oval by using skin color that is created based on a pixel value of the at least one detected face region and by displaying skin color of the face region in a color area;

extracting a skin pixel including a skin color from the received image using the skin color model, wherein when color of a pixel is classified as a skin color in the skin color model, the pixel is determined to be a pixel of the skin region; and

converting the extracted skin color of the extracted skin pixel into a desired skin color based on the skin color model,

wherein the multi-view is moving sub window in a screen, or using a template with diversely rotated faces, and

wherein when the received image includes at least one detected face region, the received image is a whole image including pixels inside and outside of the at least one face region.