Color segmentation

• US 8,055,067 B2
• Filed: 01/18/2007
• Issued: 11/08/2011
• Est. Priority Date: 01/18/2007
• Status: Active Grant

• Current Assignee: FotoNation Limited (Adeia, Inc.)
• Original Assignee: DigitalOptics Corporation Europe Limited (Adeia, Inc.)
• Inventors: Drimbarean, Alex, Petrescu, Stefan, Gangea, Mihnea, Bigioi, Petronel, Pososin, Alexei
• Primary Examiner(s): TUCKER, WESLEY J
• Application Number:
  US11/624,683
• Publication Number:
  US 20080175481A1
• Time in Patent Office: 1,755 Days
• Field of Search: 382/115, 382/118, 382/162, 382/164, 382/173, 382/180
• US Class Current: 382/164
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06V 40/162 using pixel segmentation or colour matching

Abstract

A method of skin segmentation of a digital image is operable in an acquisition device. An image is acquired. A value indicative of a redness of a pixel of said image is compared with a face skin pixel redness criterion. The pixel is identified as a face skin pixel if said criterion is satisfied.

Claims

1. A method of skin segmentation of a digital image, said method operable in an image acquisition device, comprising:

acquiring an image, including focusing an object through a lens onto an image sensor of an image acquisition device;

using a processor to compare

a value indicative of a redness of a pixel of said image with a face skin pixel redness criterion; and

to identify

said pixel as a face skin pixel if said criterion is satisfied, and

wherein said image is in RGB colour space, and wherein said value comprises the R value of said pixel, and

wherein said criterion comprises R>

G+K and R>

B+K, and

wherein said criterion further comprises that an intensity of said pixel, I, is greater than 240.

2. A method as claimed in claim 1,

where K is a constant approximately equal to 15 or higher.

9. A method as claimed in any of claim 2, 3, 1, 5, 7 or 8, wherein the redness criterion is proportional to the saturation of the pixel.

10. The method of any of claim 2, 3, 1, 5, 7 or 8, further comprising partitioning said image into one or more regions of similar colors and wherein said value indicative of a redness of a pixel comprises an average pixel value for one of said one or more regions.

11. The method of any of claim 2, 3, 1, 5, 7 or 8, further comprising partitioning said image into one or more regions of pixels having been identified as face skin pixels.

12. The method of any of claim 2, 3, 1, 5, 7 or 8, wherein said comparing provides a real value and wherein said method comprises providing a contour map of a region comprising said real value.

13. The method of claim 12 further comprising analyzing said contour map to locate one or more facial features within a region of face skin pixels.

14. The method of claim 12 further comprising analyzing said image to determine one or more candidate face regions and wherein said region that comprises said real value comprises one of said candidate face regions.

15. The method of claim 10 further comprising performing face detection on one or more regions of said image, each including a pixel having been identified as a face skin pixel.

16. The method of claim 11, further comprising performing face detection on one or more regions of said image, each including one of said one or more regions of face skin pixels.

17. A non-transitory, tangible computer program product comprising computer program code which when executed on a computing device is arranged to perform the method of any of claim 2-1 or 5-8.

18. A digital image processing device comprising a controller arranged to perform the method of any of claim 2-1 or 5-8.

3. A method of skin segmentation of a digital image, said method operable in an image acquisition device, comprising:

acquiring an image, including focusing an object through a lens onto an image sensor of an image acquisition device;

using a processor to compare

a value indicative of a redness of a pixel of said image with a face skin pixel redness criterion and to identify

said pixel as a face skin pixel if said criterion is satisfied, and

wherein said image is in RGB colour space, and wherein said value comprises the R value of said pixel, and

wherein said criterion comprises R>

G+K and R>

B+K, where K comprises approximately

x*(pixel saturation+y)

where x is in the range of approximately 0.53 to 0.6; and

y is in the range of approximately 5 to 6.5.

4. The method of claim 3 comprising calculating said pixel saturation as:

Sqrt(R²*0.27847−

R*G*0.30610+G²*0.28503−

R*B*0.25005+B²*0.25661−

G*B*0.26317).

5. A method of skin segmentation of a digital image, said method operable in an image acquisition device, comprising:

acquiring an image, including focusing an object through a lens onto an image sensor of an image acquisition device;

using a processor to compare

a value indicative of a redness of a pixel of said image with a face skin pixel redness criterion and to identify

said pixel as a face skin pixel if said criterion is satisfied, and

wherein said image is in YCC colour space, and wherein said value is a Cr value of said pixel, and

wherein said criterion comprises;

Cr>

148.8162−

0.1626*Cb+0.4726*K and

Cr>

1.2639*Cb−

33.7803+0.133*K.

6. The method of claim 5 where K is a constant approximately equal to 15 or higher.

7. A method of skin segmentation of a digital image, said method operable in an image acquisition device, comprising:

acquiring an image, including focusing an object through a lens onto an image sensor of an image acquisition device;

using a processor to compare

a value indicative of a redness of a pixel of said image with a face skin pixel redness criterion and to identify

said pixel as a face skin pixel if said criterion is satisfied, and

wherein said image is in YCC colour space, and wherein said value is a Cr value of said pixel, and

wherein said criterion comprises;

Cr>

148.8162−

0.1626*Cb+0.2836*(pixel saturation+5) and

Cr>

1.2639*Cb−

33.7803+0.4279*(pixel saturation+5)

where pixel saturation is;

√

{square root over (((Cr−

128)²+(Cb−

128)²))}{square root over (((Cr−

128)²+(Cb−

128)²))}.

8. A method of skin segmentation of a digital image, said method operable in an image acquisition device, comprising:

acquiring an image, including focusing an object through a lens onto an image sensor of an image acquisition device;

using a processor to compare

a value indicative of a redness of a pixel of said image with a face skin pixel redness criterion and to identify

said pixel as a face skin pixel if said criterion is satisfied, and

wherein said image is in YCC colour space, and wherein said value is a Cr value of said pixel, and

wherein said criterion further comprises an alternate condition that an intensity of said pixel, Y, is greater than 240.