Granted Patent B2

US 7,433,085 · App. 10/414,530 · Granted Oct 7, 2008

Abstract

Claims (45)

1. A method for processing color pixel data of an image, comprising the steps of:

computing in hardware logic using bit-shift operations an approximation of at least one of a plurality of other color space components of a pixel represented in RGB color space, wherein the at least one other color space component includes luminance, and an approximation of luminance (Y) is computed by summing a red component (R) of the pixel shifted right two bits, a green component (G) of the pixel shifted fight by one bit, and a blue component (B) of the pixel shifted right two bits; and

evaluating a characteristic of the pixel in response to the computed approximation of the at least one of the plurality of other color space components.

2. The method claimed in claim 1 , wherein the step of evaluating a characteristic of the pixel comprises computing a histogram from one of the computed other color space components of the pixels.

3. The method claimed in claim 2 , wherein the step of computing a histogram comprises the steps of:

defining a plurality of luminance intervals each spanning a plurality of luminance values;

determining a frequency of occurrence of pixels in the image having a luminance within each luminance interval;

identifying a luminance interval with the highest frequency of occurrence; and

performing an image processing method on pixels of the image having a luminance related by a predetermined relation to a luminance value in the identified luminance interval.

4. The method claimed in claim 3 , wherein:

the predetermined relation is that pixels on which the image processing method is performed have a luminance greater than a cutoff value (Cutoff) that is a predetermined percentage of the luminance value in the identified luminance interval; and

the image processing method is background-removal and comprises the step of setting to a predetermined maximum or white value the red, green and blue components of each pixel in the image having a luminance greater than or equal to Cutoff.

5. The method claimed in claim 4 , wherein the predetermined percentage is approximately 80%.

6. The method claimed in claim 4 , wherein the background-removal method further comprises the step of expanding the red, green and blue components of each pixel in the image having a luminance value less than Cutoff.

7. The method claimed in claim 6 , wherein expanding the red, green and blue components of each pixel comprises the step of multiplying each of the red, green and blue components by 255 and dividing by Cutoff.

8. The method claimed in claim 1 , wherein the step of evaluating a characteristic of the pixel comprises comparing at least one of the computed other color space components of the pixels to a predetermined threshold value.

9. The method claimed in claim 8 , wherein computing at least one of a plurality of other color space components of a pixel represented in RGB color space further comprises the steps of:

computing an approximation of chrominance component Cr of the pixel by computing ((R>>1)+(R>>2)−(R>>5))−((Y>>1)+(Y>>2)−(Y>>5)), wherein “X>>n” represents shifting RGB color space component X by n bits to the right; and

computing an approximation of chrominance component Cb of the pixel by computing ((B>>1)+(B>>4))−((Y>>1)+(Y>>4)).

10. The method claimed in claim 9 , wherein comparing at least one of the computed other color space components of the pixels to a predetermined threshold value comprises the steps of:

determining whether the computed approximation of Cr is less than a predetermined Cr upper threshold;

determining whether the computed approximation of Cr is greater than a predetermined Cr lower threshold;

determining whether the computed approximation of Cb is less than a predetermined Cb upper threshold;

determining whether the computed approximation of Cb is greater than a predetermined Cb lower threshold;

determining whether the computed approximation of Y is greater than a predetermined Y cutoff; and

determining whether the computed approximation of Y is less than a predetermined Y lower threshold.

11. The method claimed in claim 10 , wherein evaluating a characteristic of the pixel further comprises the steps of:

setting the pixel to white if the computed approximation of Cr is less than the predetermined Cr upper threshold, and the computed approximation of Cr is greater than the predetermined Cr lower threshold, and the computed approximation of Cb is less than the predetermined Cb upper threshold, and the computed approximation of Cb is greater than the predetermined Cb lower threshold, and the computed approximation of Y is greater than the predetermined Y cutoff; and

setting the pixel to black if the computed approximation of Cr is not less than the predetermined Cr upper threshold, or the computed approximation of Cr is not greater than the predetermined Cr lower threshold, or the computed approximation of Cb is not less than the predetermined Cb upper threshold, or the computed approximation of Cb is not greater than the predetermined Cb lower threshold, or the computed approximation of Y is not greater than the predetermined Y cutoff, and if the computed approximation of Y is less than the predetermined Y lower threshold.

12. A system for processing color pixel data of an image, comprising:

color space converter logic having bit-shift logic producing approximations of YCbCr color space components of a pixel represented in RGB color space; and

a histogram generator producing a histogram from one of the computed other color space components of the pixels, wherein the histogram generator comprises:

a plurality of accumulators, each defining exactly one luminance interval spanning a plurality of luminance values;

first comparator logic determining a frequency of occurrence of pixels in the image having a luminance within each luminance interval;

second comparator logic identifying a luminance interval with the highest frequency of occurrence; and

background removal logic identifying and setting to a predetermined maximum or white value the red, green and blue components of each pixel in the image having a luminance greater than or equal to a cutoff value (Cutoff) that is a predetermined percentage of the luminance value in the identified luminance interval, wherein the predetermined percentage is approximately 80%.

13. A system for processing color pixel data of an image, comprising:

color space converter logic having bit-shift logic producing approximations of YCbCr color space components of a pixel represented in RGB color space; and

a histogram generator producing a histogram from one of the computed other color space components of the pixels, wherein the histogram generator comprises:

a plurality of accumulators, each defining exactly one luminance interval spanning a plurality of luminance values;

first comparator logic determining a frequency of occurrence of pixels in the image having a luminance within each luminance interval;

second comparator logic identifying a luminance interval with the highest frequency of occurrence;

background removal logic identifying and setting to a predetermined maximum or white value the red, green and blue components of each pixel in the image having a luminance greater than or equal to a cutoff value (Cutoff) that is a predetermined percentage of the luminance value in the identified luminance interval; and

pixel expansion logic expanding the red, green and blue components of each pixel in the image having a luminance value less than Cutoff.

14. The system claimed in claim 13 , wherein the pixel expansion logic multiplies each of the red, green and blue components by 255 and divides by Cutoff.

Assignments (8)