System and method for adjusting color data of pixels in a digital image

• US 6,204,858 B1
• Filed: 05/30/1997
• Issued: 03/20/2001
• Est. Priority Date: 05/30/1997
• Status: Expired due to Term

• Current Assignee: Adobe Systems Incorporated (Adobe Inc.)
• Original Assignee: Adobe Systems Incorporated (Adobe Inc.)
• Inventors: Gupta, Naresh C.
• Primary Examiner(s): Zimmerman, Mark
• Assistant Examiner(s): MANCUSO, HUEDUNG XUAN CAO
• Application Number:
  US08/866,093
• Time in Patent Office: 1,390 Days
• Field of Search: 345/431, 382/163, 382/167, 358/518, 358/519, 358/520, 358/521, 358/523
• US Class Current: 345/600
• CPC Class Codes: G06T 2207/30216 Redeye defect
• CPC Class Codes: G06T 7/90 Determination of colour characteristics
• CPC Class Codes: H04N 1/62 Retouching, i.e. modification of isolated colours only or in isolated picture areas only
• CPC Class Codes: H04N 1/624 Red-eye correction control of camera based on recognised facial parts H04N23/611

Abstract

Color data of pixels of a digital image are adjusted by identifying pixels of a digital image having original color data corresponding to predetermined color and shape characteristics and adjusting the original color data of the identified pixels to achieve a desired result.

Claims

1. A method for adjusting color values of pixels of an image to reduce a red-eye effect, the method comprising:

generating a red-enhanced value for each pixel in the image, wherein the red-enhanced value of a pixel represents the degree of redness of the pixel;

convolving a plurality of spatial filters with the red-enhanced values to create one or more red filter values, each of the one or more red filter values corresponding to a pixel region having a specific shape and size;

identifying a pixel region having a shape and size defined by at least one of the plurality of spatial filters as a red-eye pixel region if the corresponding red filter value exceeds a predetermined threshold; and

adjusting the color values of the red-eye pixel region to reduce the red-eye effect.

2. The method of claim 1, wherein generating a red-enhanced value for each pixel comprises:

categorizing each pixel in the image as a non-red pixel or a red pixel;

assigning a minimum red-enhanced value to each non-red pixel; and

calculating a red-enhanced value for each red pixel, wherein the red-enhanced value represents a degree of redness of the pixel.

3. The method of claim 2, wherein the color values of the pixels of the image comprise the values for red, green and blue components, and wherein:

each pixel having a red component value exceeding both the green component value and the blue component value by first and second predetermined thresholds, respectively, is categorized as a red pixel; and

each remaining pixel is categorized a non-red pixel.

4. The method of claim 1, wherein the image is a portion of an original image.

5. The method of claim 4, wherein the image is a user-selected portion of the original image.

6. The method of claim 1, wherein adjusting the color values comprises:

identifying pixels in the image corresponding to an iris area of an eye affected by the red-eye effect;

determining a color value of the iris pixels; and

adjusting the color value of each red-eye pixel to approach the color value of the iris pixels.

7. The method of claim 6, wherein determining the color value of the iris pixels comprises using a color value of a default color.

8. The method of claim 7, wherein the default color is selected by a user.

9. A method for identifying a red-eye area in an image, comprising:

generating a red-enhanced value for each pixel in the image;

generating a weak-color-enhanced value for each pixel in the image;

convolving an inner filter with the red-enhanced values to create a red filter value corresponding to a pixel region having a specific shape and size defined by the inner filter;

convolving an outer spatial filter with the weak-color-enhanced values to create a weak-color filter value corresponding to a pixel region having a specific shape and size defined by the outer filter; and

identifying a red-eye area having a shape and size defined by at least one of the inner filter and the outer filter using both the red-enhanced values and the weak-color-enhanced values of the pixels.

10. The method of claim 9, wherein identifying a red-eye area comprises:

identifying as a red-eye area an area of the image having a ring-shaped region of weak-color-background pixels surrounding a circular-shaped region of red-eye pixels.

11. The method of claim 10, further comprising:

convolving a plurality of ring-shaped outer filters with the weak-color-enhanced values to create one or more ring filter values, each of the one or more ring filter values corresponding to a pixel region having a specific size and shape;

convolving a plurality of circular-shaped inner filters with the red-enhanced values to create one or more red filter values, each of the one or more red filter values corresponding to a pixel region having a specific size and shape; and

using the ring filter values and the red filter values to identify a red-eye area in the image.

12. A computer program product stored on a machine-readable medium for adjusting color values of pixels of an image to reduce a red-eye effect, the product comprising instructions operable to cause a programmable processor to:

generate a red-enhanced value for each pixel in the image, wherein the red-enhanced value of a pixel represents the degree of redness of the pixel;

convolve a plurality of spatial filters with the red-enhanced values to create one or more red filter values, each of the one or more red filter values corresponding to a pixel region having a specific shape and size;

identify a pixel region having a shape and size defined by at least one of the plurality of spatial filters as a red-eye pixel region if the corresponding red filter value exceeds a predetermined threshold; and

adjusting the color values of the red-eye region to reduce the red-eye effect.

13. The product of claim 12, wherein generating a red-enhanced value for each pixel comprises:

categorizing each pixel in the image as a non-red pixel or a red pixel;

assigning a minimum red-enhanced value to each non-red pixel; and

calculating a red-enhanced value for each red pixel, wherein the red-enhanced value represents a degree of redness of the pixel.

14. The product of claim 13, wherein the color values comprise values for red, green and blue components, and wherein:

each pixel having a red component value exceeding both the green component value and the blue component value by first and second predetermined thresholds, respectively, is categorized as a red pixel; and

each remaining pixel is categorized a non-red pixel.

15. The product of claim 12, wherein the image is a portion of an original image.

16. The product of claim 15, wherein the image is a user-selected portion of the original image.

17. The product of claim 12, wherein adjusting the color values comprises:

identifying pixels in the image corresponding to an iris area of an eye affected by the red-eye effect;

determining a color value of the iris pixels; and

adjusting the color value of each red-eye pixel to approach the color value of the iris pixels.

18. The product of claim 17, wherein determining the color value of the iris pixels comprises using a color value of a default color.

19. The product of claim 18, wherein the default color is selected by a user.

20. A computer program product stored on a machine-readable medium for identifying a red-eye area in an image, the product comprising instructions operable to cause a programmable processor to:

generate a red-enhanced value for each pixel in the image;

generate a weak-color-enhanced value for each pixel in the image;

convolve an inner filter with the red-enhanced values to create a red filter value corresponding to a pixel region having a specific shape and size defined by the inner filter;

convolve an outer spatial filter with the weak-color-enhanced values to create a weak-color filter value corresponding to a pixel region having a specific shape and size defined by the outer filter; and

identify a red-eye area having a shape and size defined by at least one of the inner filter and the outer filter using both the red-enhanced values and the weak-color-enhanced values of the pixels.

21. The product of claim 20, wherein identifying a red-eye area comprises:

identifying as a red-eye area an area of the image having a ring-shaped region of weak-color-background pixels surrounding a circular-shaped region of red-eye pixels.

22. The product of claim 21, further comprising instructions operable to cause a programmable processor to:

convolve a plurality of ring-shaped outer filters with the weak-color-enhanced values to create one or more ring filter values, each of the one or more ring filter values corresponding to a pixel region having a specific size and shape;

convolve a plurality of circular-shaped inner filters with the red-enhanced values to create one or more red filter values, each of the one or more red filter values corresponding to a pixel region having a specific size and shape; and

use the ring filter values and the red filter values to identify a red-eye area in the image.