Method for representing a digital color image using a set of palette colors based on detected important colors
View Patent ↗Disclosed is a method for converting an input digital color image having a set of possible input colors to an output digital color image having a set of palette colors, the number of palette colors being less than the number of possible input colors, wherein the set of palette colors is determined based on the distribution of colors in the input digital image boosted by a distribution of important colors contained in the input digital image.
1. A method for converting an input digital color image having a set of possible input colors to an output digital color image having a set of palette colors, the number of palette colors being less than the number of possible input colors, wherein the set of palette colors is determined based on the distribution of colors in the input digital image supplemented by a distribution of pre-determined important colors, comprising:
a) determining a distribution of input colors using pixels in the input digital color image;
b) detecting any regions of predetermined important colors in the input digital color image;
c) collecting additional pixels from any detected regions of important colors;
d) adding the collected additional pixels to the distribution of input colors to create a boosted distribution of colors;
e) determining a set of palette colors to be used in the formation of an output digital color image in response to the boosted distribution of colors; and
f) forming the output digital color image by assigning each color in the input digital color image to one of the colors in the set of palette colors.
2. The method of claim 1 , wherein the regions of the predetermined important colors includes regions of skin-tone colors.
3. The method of claim 1 , wherein the regions of the predetermined important colors include regions of neutral colors.
4. The method of claim 1 , wherein the regions of the predetermined important colors include regions of sky colors.
5. The method of claim 1 , wherein determining the boosted distribution of colors is accomplished by appending additional pixels to the input digital color image to form an enlarged input digital color image, where the additional pixels are determined from the detected regions of the predetermined important colors in the input digital color image, and then determining the distribution of colors in the enlarged input digital color image.
6. The method of claim 1 , wherein the set of palette colors is determined using a sequential scalar quantization algorithm.
7. The method of claim 6 , wherein the sequential scalar quantization algorithm includes:
i) sequentially partitioning the colors of the boosted distribution of colors into a set of color space regions; and
ii) determining the set of palette colors by selecting an output color for each color space region in the set of color space regions.
8. The method of claim 7 , comprising:
further determining the color value for each pixel of the output digital color image by identifying the palette color corresponding to the color space region containing the input color for the corresponding pixel of the input digital color image.
9. The method of claim 1 , wherein the set of palette colors is determined using a vector quantization algorithm.
10. The method of claim 1 , wherein the output digital color image is formed by assigning each color in the input digital color image to the color in the set of palette colors having the smallest color difference relative to the color of the input digital color image.
11. The method of claim 1 , wherein step e) includes the use of a multi-level halftoning technique to assign each color in the input digital color image to one of the colors in the set of palette colors to approximately preserve the local mean color value.
12. The method of claim 11 , wherein the multi-level halftoning technique is an error diffusion technique that distributes the quantization errors introduced when processing an input pixel to nearby input pixels that have not yet been processed.
13. The method of claim 1 , wherein the detecting regions of the predetermined important colors step includes
i) obtaining probability values for the predetermined important colors in the input digital color image; and
ii) extracting regions of the predetermined important colors consisting pixels whose probability values are above a predetermined threshold.
14. The method of claim 13 , wherein the probability values for the predetermined important colors are probability values for skin-tone colors.
15. The method of claim 13 , further comprising:
applying a scene balance algorithm to the input color digital image before obtaining the probability values.
16. The method of claim 1 , wherein detecting the regions of the predetermined important colors corresponds to detecting regions of typical skin-tone colors, and the detecting step further includes:
i) detecting human faces in the input color digital image; and
ii) extracting regions of skin-tone colors from the detected face regions.
17. A method for converting an input digital color image having a set of possible input colors to an output digital color image having a set of palette colors, the number of palette colors being less than the number possible input colors, wherein the set of palette colors is determined based on the distribution of colors in the first digital image boosted by a distribution of the predetermined important colors contained in the input digital image, comprising:
a) appending additional pixels to the input digital color image to form an enlarged input digital color image, where the additional pixels are collected from any detected regions of the predetermined important colors in the input digital color image;
b) determining the distribution of colors in the enlarged input digital color image;
c) determining the set of palette colors to be used in the formation of the output digital color image responsive to the distribution of colors in the enlarged input digital color image; and
d) forming the output digital color image by assigning each color in the input digital color image to one of the colors in the set of palette colors.
18. The method of claim 17 , where the additional pixels are provided in the form of a target image containing color patches extracted from the detected regions of the predetermined important colors in the input color digital image.
19. The method of claim 18 , where the target image is resized to match the size of the input digital color image.
20. A computer storage medium having instructions stored therein for causing the computer to convert an input digital color image having a set of possible input colors to an output digital color image having a set of palette colors, the number of palette colors being less than the number of possible input colors, wherein the set of palette colors is determined based on the distribution of colors in the input digital image supplemented by a distribution of pre-determined important colors including
a) determining a distribution of input colors using pixels in the input digital color image;
b) detecting any regions of predetermined important colors in the input digital color image;
c) collecting additional pixels from any detected regions of important colors;
d) adding the collected additional pixels to the distribution of input colors to create a boosted distribution of colors;
e) determining a set of palette colors to be used in the formation of an output digital color image in response to the boosted distribution of colors; and
f) forming the output digital color image by assigning each color in the input digital color image to one of the colors in the set of palette colors.