# US 2010/0166304 A1 · Method for Adjusting Skin Color of Digital Image

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2010/0166304 A1
- Document ID: US20100166304A1
- Status: Published application
- Kind: A1
- Application number: 12/428,920
- Filed: 2009-04-23
- Published: 2010-07-01
- Earliest claimed: 2008-12-31
- Assignees: Altek Corporation
- Inventors: Chan Chen-Hung; Chou Hong-Long
- Web page: https://patentgrep.com/patent/US20100166304A1

## CPC

- H04N 1/628: {Memory colours, e.g. skin or sky}
- G06T 7/11: Region-based segmentation
- G06T 7/90: Determination of colour characteristics
- G06V 10/56: relating to colour
- G06T 2207/10024: Color image
- G06T 2207/30201: Face
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A method for adjusting a skin color of a digital image adjusts the skin color of an input image. The method includes performing a skin color detection process on the input image to generate a skin-color probability plot Sp in a size corresponding to the input image; providing a hue-saturation lookup table named LUT_Color; performing a skin-color reproduction process on the input image to look up the LUT_Color for a chrominance pixel value for each pixel value of the input image to generate a first image, and adjust each pixel value of the first image by using the skin-color probability plot Sp to generate a second image; performing a skin color smoothing process on the second image to generate a third image; and mixing pixel values of the input image and the third image to generate a target image.

## Claims

### Claim 1

A method for adjusting a skin color of a digital image, adapted to adjust the skin color of an input image, the method comprising:
loading the input image;
performing a skin color detection process on the input image to generate a skin-color probability plot Sp in a size corresponding to the input image;
providing a hue-saturation lookup table named LUT_Color;
performing a skin color reproduction process on the input image to look up the LUT_Color for a chrominance pixel value of a corresponding color for each pixel value of the input image to generate a first image, and adjust each pixel value of the first image by using the skin-color probability plot Sp to generate a second image;
performing a skin color smoothing process on the second image to generate a third image; and
performing an image mixing process to mix pixel values of the input image and the third image at a corresponding mixing ratio to generate a target image.

### Claim 2 (depends on 1)

The method for adjusting a skin color of a digital image according to claim 1, wherein the input image at least comprises a luminance image Y, a blue chrominance image Cb, and a red chrominance image Cr.

### Claim 3 (depends on 2)

The method for adjusting a skin color of a digital image according to claim 2, wherein the step of looking up the LUT_Color further comprises:
selecting a pixel from the input image;
obtaining a blue chrominance pixel value and a red chrominance pixel value from the blue chrominance image Cb and the red chrominance image Cr respectively according to a position of the selected pixel;
for the blue chrominance image, looking up the LUT_Color according to the blue chrominance pixel value and the red chrominance pixel value to generate a first blue chrominance image Cba, wherein
Cba=LUT_Color [Cb] [Cr]
for the red chrominance image, looking up the LUT_Color according to the blue chrominance pixel value and the red chrominance pixel value to generate a first red chrominance image Cra, wherein
Cra=LUT_Color [Cr] [Cb]
and
generating the first image by using the luminance image Y, the first blue chrominance image Cba and the first red chrominance image Cra.

### Claim 4 (depends on 1)

The method for adjusting a skin color of a digital image according to claim 1, wherein the step of adjusting each pixel value of the first image by using the skin-color probability plot Sp further comprises:
obtaining a first blue chrominance image Cba and a first red chrominance image Cra;
respectively performing a calculation at a corresponding ratio on each pixel in the first blue chrominance image Cba and the blue chrominance image Cb according to the skin-color probability plot Sp, to output a second blue chrominance image Cbc, wherein
Cbc(x,y)=Sp(x,y)×Cba(x,y)+(255−Sp(x,y)×Cb(x,y))
and (x,y) is position coordinates of a pixel;
respectively performing a calculation at a corresponding ratio on each pixel in the first red chrominance image Cra and the red chrominance image Cr according to the skin-color probability plot Sp, to output a second red chrominance image Crc, wherein
Crc(x,y)=Sp(x,y)×Cra(x,y)+(255−Sp(x,y)×Cr(x,y))
and (x,y) is position coordinates of a pixel; and
generating the second image according to the second blue chrominance image Cbc and the second red chrominance image Crc.

### Claim 5 (depends on 1)

The method for adjusting a skin color of a digital image according to claim 1, wherein the step of performing the skin color smoothing process further comprises:
setting a selection window for selecting areas at a same position and of a same pixel array size in the images; and
respectively selecting image areas from the luminance image Y, a second blue chrominance image Cbc, and a second red chrominance image Crc by using the selection window, and processing the selected image areas by a filter, to respectively generate a second luminance image YLP, a third blue chrominance image CbLP, and a third red chrominance image CrLP in the third image, wherein
Y
LP

(
x
,
y
)
=
∑
Y

(
i
,
j
)
N
Cb
LP

(
x
,
y
)
=
∑
Cb
c

(
i
,
j
)
N
Cr
LP

(
x
,
y
)
=
∑
Cr
c

(
i
,
j
)
N
(x,y) is position coordinates of a pixel, and N is a number of pixels in the selection window.

### Claim 6 (depends on 5)

The method for adjusting a skin color of a digital image according to claim 5, wherein the filter is a low-pass filter or a nonlinear filter.

### Claim 7 (depends on 5)

The method for adjusting a skin color of a digital image according to claim 5, wherein the step of performing the image mixing process further comprises:
providing a mixing ratio α;
mixing pixel values at corresponding positions in the luminance image Y and the second luminance image YLP at the mixing ratio α to generate a target luminance image Yo, wherein
Yo(x,y)=α×Y(x,y)+(1−α)×YLP(x,y))
mixing pixel values at corresponding positions in the first blue chrominance image Cba and the third blue chrominance image CbLP at the mixing ratio α to generate a target blue chrominance image Cbo, wherein
Cbo(x,y)=α×Cb(x,y)+(1−α)×CbLP(x,y))
mixing pixel values at corresponding positions in the first red chrominance image Cra and the third red chrominance image CrLP at the mixing ratio α to generate a target red chrominance image Cro, wherein
Cro(x,y)=α×Cr(x,y)+(1−α)×CrLP(x,y))
and
generating the target image according to the target luminance image Yo, the target blue chrominance image Cbo, and the target red chrominance image Cro.