Patent US6690822B1 successfully refreshed from Google Patents.
US6690822B1
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method for detecting skin color in a digital image
Assignee: Intellectual Ventures Fund 83 LLC
•
Filing: 2000-10-20
•
Granted: 2004-02-10
•
Expires: 2020-10-20
Abstract
Abstract A method for detecting skin color in a digital image having pixels in an RGB color space generally includes the steps of performing statistical analysis of the digital color image to determine the mean RGB color values; then, if the mean value of any one of the colors is below a predetermined threshold, applying a transformation to the digital image to move skin colors in the image toward a predetermined region of the color space; and employing the transformed space to locate the skin color pixels in the digital color image. More specifically, if the mean value of any one of the colors is below a predetermined threshold, a non-linear transformation is applied to the digital image to move skin colors in the image toward a predetermined region of the color space. Then, depending on the preceding step, either the digital image or the transformed digital image is converted from the RGB space to a generalized RGB space to produce a gRGB digital image; skin color pixels are detected within the gRGB digital image; a first skin color image mask is formed based on the detected skin color pixels; a masked gRGB image is generated using the first skin color image mask; and finally the skin color image mask is employed to locate the skin color pixels in the digital color image.
Inventors
- 👤 Shoupu Chen
- 👤 Lawrence A. Ray
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 14:05
CPC Classifications
G; G06; G06T; G06T7/00; G06T7/10; G06T7/11; G06V; G06V10/00; G06V10/40; G06V10/56; G06V40/00; G06V40/10; G06V40/16; G06V40/161; G06V40/162; G06T2207/00; G06T2207/30; G06T2207/30196; G06T2207/30201
Claims Analysis
Total Claims: 17
Claim 1
Independent Claim
243 words
A method for detecting skin color in a digital image having pixels in an RGB color space, comprising the steps of: a) performing statistical analysis of the digital color image to determine the mean RGB color values; b) if the mean value of any one of the colors is below a predetermined threshold, applying a transformation to the digital image to move skin colors in the image toward a predetermined region of the color space; c) converting the digital image, or the transformed digital image from step (b), from the RGB space to a generalized RGB space, to produce a gRGB digital image; d) detecting skin color pixels within the gRGB digital image; e) forming a first skin color image mask based on the detected skin color pixels; f) generating a masked gRGB image using the first skin color image mask; g) converting the masked gRGB image to a hue image; h) removing possible blond hair color pixels from the hue image to produce a modified hue image; i) forming a second skin color image mask based on the skin color pixels in the modified hue image; j) if the second skin color image mask is smaller than the first skin color image mask by a predetermined amount, selecting the first skin color image mask, otherwise, selecting the second skin color image mask; and k) employing the selected skin color image mask to locate the skin color pixels in the digital color image.
Claim 2
Independent Claim
243 words
A method for detecting skin color in a digital image having pixels in an RGB color space, comprising the steps of: a) performing statistical analysis of the digital color image to determine the mean RGB color values; b) if the mean value of any one of the colors is below a predetermined threshold, applying a transformation to the digital image to move skin colors in the image toward a predetermined region of the color space; c) converting the digital image, or the transformed digital image from step (b), from the RGB space to a generalized RGB space, to produce a gRGB digital image; d) detecting skin color pixels within the gRGB digital image; e) forming a first skin color image mask based on the detected skin color pixels; f) generating a masked gRGB image using the first skin color image mask; g) converting the masked gRGB image to a hue image; h) removing possible blond hair color pixels from the hue image to produce a modified hue image; i) forming a second skin color image mask based on the skin color pixels in the modified hue image; j) if the second skin color image mask is smaller than the first skin color image mask by a predetermined amount, selecting the first skin color image mask, otherwise, selecting the second skin color image mask; and k) employing the selected skin color image mask to locate the skin color pixels in the digital color image.
Claim 3
Dependent Claim
60 words
A method as claimed in claim 1 wherein the step (a) of performing statistical analysis of the input digital color (RGB) image comprising the steps of: a1) rank-order filtering the input digital image to remove near white or saturated color pixels; and a2) computing a mean value for each individual (RGB) color channel of the rank-order filtered digital RGB image.
Claim 4
Dependent Claim
87 words
A method as claimed in claim 2 , wherein the step (b) of applying a transformation to the input digital color (RGB) image comprises the steps of: b2) making a decision whether a non-linear transformation is necessary for the input digital RGB image based on the mean statistics computed in step (a2) for each individual (RGB) color channel of the rank-order filtered digital RGB image, and b2) based on the decision in step (b1), applying or not applying the non-linear transformation to the input digital RGB image.
Claim 5
Dependent Claim
37 words
A method as claimed in claim 3 wherein the step (b2) of applying non-linear transformation to the input digital RGB image comprises transforming the input digital RGB image to another digital RGB image through color histogram equalization.
Claim 6
Dependent Claim
54 words
A method as claimed in claim 1 wherein step (c) of converting the input digital image from the RGB space to a generalized RGB space comprises the steps of: c1) accepting a digital RGB image either before or after the non-linear transformation, and c2) converting the digital RGB image to a digital gRGB image.
Claim 7
Dependent Claim
129 words
A method as claimed in claim 1 wherein step (d) of detecting skin color pixels within the gRGB digital image comprises steps of: d1) projecting the gRGB digital image to a gRG space; d2) performing skin color classification in the gRG space using a first ellipse classifier with a first ellipse vector; d3) evaluating the skin classification result in step (d2); d4) performing skin color classification in the gRG space using a second ellipse classifier with a second ellipse vector based on the evaluation results in step (d3); d5) evaluating the results from the first and second ellipse classifiers; d6) forming a skin color region mask based on the steps of (d3) and (d5), and d7) generating a masked digital gRGB image using the mask formed in step (d6).
Claim 8
Dependent Claim
25 words
A method as claimed in claim 6 wherein the ellipse classifiers of steps d2) and d4) are trained on a plurality of digital RGB images.
Claim 9
Dependent Claim
92 words
A method as claimed in claim 7 wherein the training of the ellipse classifier comprises steps of: randomly selecting a plurality of digital RGB images from a larger digital RGB image pool; manually selecting skin color RGB patches from the selected digital RGB images; converting the selected skin color RGB patches to generalized RGB space, obtaining gRGB skin color patches; projecting the gRGB skin color patches onto a gRG space; approximating the skin color region in the gRG space with an ellipse, and estimating the ellipse parameters and obtaining an ellipse vector.
Claim 10
Dependent Claim
27 words
A method as claimed in claim 1 , wherein the step (h) of the removal of blond hair color pixels is performed in one dimensional hue space.
Claim 11
Dependent Claim
14 words
A computer program product for performing the method as claimed in claim 1 .
Claim 12
Independent Claim
92 words
A method for detecting skin color in a digital image having pixels in an RGB color space, comprising the steps of: a) performing statistical analysis of the digital color image to determine the mean RGB color values; b) if the mean value of any one of the colors is below a predetermined threshold, applying a transformation to the digital image to move skin colors in the image toward a predetermined region of the color space; and c) employing the transformed space to locate the skin color pixels in the digital color image.
Claim 13
Independent Claim
92 words
A method for detecting skin color in a digital image having pixels in an RGB color space, comprising the steps of: a) performing statistical analysis of the digital color image to determine the mean RGB color values; b) if the mean value of any one of the colors is below a predetermined threshold, applying a transformation to the digital image to move skin colors in the image toward a predetermined region of the color space; and c) employing the transformed space to locate the skin color pixels in the digital color image.
Claim 14
Dependent Claim
77 words
The method as claimed in claim 11 wherein step (c) further comprises the steps of: converting the transformed digital image from the RGB space to a generalized RGB space, to produce a gRGB digital image; detecting skin color pixels within the gRGB digital image; forming a first skin color image mask based on the detected skin color pixels; and employing the first skin color image mask to locate the skin color pixels in the digital color image.
Claim 15
Dependent Claim
64 words
The method as claimed in claim 12 wherein step (c) further comprises: forming a second skin color image mask based on the detected skin color pixels; selecting either the first skin color image mask or the second skin color image mask based on predetermined criteria; and employing the selected skin color image mask to locate the skin color pixels in the digital color image.
Claim 16
Dependent Claim
124 words
The method as claimed in claim 12 wherein step (c) further comprises the steps of: generating a masked gRGB image using the first skin color image mask; converting the masked gRGB image to a hue image; removing possible blond hair color pixels from the hue image to produce a modified hue image; forming a second skin color image mask based on the skin color pixels in the modified hue image; if the second skin color image mask is smaller than the first skin color image mask by a predetermined amount, selecting the first skin color image mask, otherwise, selecting the second skin color image mask; and employing the selected skin color image mask to locate the skin color pixels in the digital color image.
Claim 17
Dependent Claim
14 words
A computer program product for performing the method as claimed in claim 11 .
References Cited (Backward Citations)
US4203671A
Cited
US5128711A
Cited
US6249317B1
Cited
US6072526A
Cited
US5585860A
Cited
US5715377A
Cited
US5940530A
Cited
US5850463A
Cited
US6343141B1
Cited
US6611613B1
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2007-02-09 | 2007-08-11 | 2008-02-10 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2011-02-09 | 2011-08-11 | 2012-02-10 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2015-02-09 | 2015-08-11 | 2016-02-10 | $7700.00 | Window Closed / Lapsed |