Detection of anti aliasing in two-color images for improved compression
A method for identifying anti aliased images of two or three colors makes use of the fact that in most anti-aliasing techniques, the color values of the source images are known with sub-pixel accuracy, e.g., the boundaries of text characters are described with arbitrary high precision and averaging ensures that the destination pixel values are interpolations of the two source colors so that all of the color values of the image lie on a single line, or a single plane, in 3-dimensional color space.
1. A method for detecting anti-aliased images comprising:
determining a threshold number of unique colors;
counting using a CPU a number of unique colors of pixels in an image, and if said number of unique colors of pixels in said image is no greater than said threshold number, determining that said image may be a two-color anti-aliased image;
determining two extremal colors of said image;
calculating using a CPU if each pixel has a color which lies on a line connecting said two extremal colors in color space, and if each pixel of said image has a color which lies on the line connecting said two extremal colors in color space, then verifying that said image is a two-color anti-aliased image; and
in response to verifying that said image is a two-color anti-aliased image, compressing data representing said image as data representing a two-color anti-aliased image to generate compressed data representing said image.
2. The method of claim 1 , further comprising transmitting said compressed data representing said image for display on a display device.
3. The method of claim 1 , wherein; said color space is RGB space.
4. The method of claim 1 , wherein; said color space is YUV space.
5. The method of claim 1 , wherein; said color space is YCbCr space.
6. The method of claim 1 , wherein the extremal colors of said image are black and white.
7. A method for detecting anti-aliased images comprising:
scanning each pixels in an image using a CPU;
calculating using a CPU if each of said pixels of said image has a color on a line R=G=B in RGB color space and, if each of said pixels has a color on the line R=G=B in RGB color space, then determining that said image is a grey-scale anti-aliased image; and
in response to determining that said image is a grey-scale anti-aliased image, compressing data representing said image as data representing a grey-scale anti-aliased image to generate compressed data representing said image.
8. The method of claim 7 , further comprising transmitting said compressed data representing said image for display on a display device.
9. A method for detecting anti-aliased images comprising:
scanning each pixel in an image using a CPU;
calculating using a CPU whether each of said pixels has a color on a line between black and white in a color space and, if each of said pixels has a color on the line between black and white in a color space, then determining that said image is a grey-scale anti-aliased image; and
in response to determining that said image is a grey-scale anti-aliased image, compressing data representing said image as data representing a grey-scale anti-aliased image.
10. The method of claim 9 , further comprising transmitting said compressed data representing said image for display on a display device.
11. The method for detecting anti-aliased images of claim 9 , wherein; said color space is RGB space and said calculating step comprises:
calculating using a CPU if each of said pixels has a color on the line R=B=G in RGB color space, and if each of said pixels has a color on the line R=B=G in RGB color space then determining that said image is a grey-scale anti-aliased image.
12. The method for detecting anti-aliasing images of claim 9 , wherein; said color space is YUV space.
13. The method for detecting anti-aliased images of claim 9 , wherein; said color space is YCbCr space.
14. The method for detecting anti-aliased images of claim 10 , wherein; said color space is RGB space and said calculating step comprises:
calculating using a CPU if each of said pixels has a color on the line R=B=G in RGB color space, and if each of said pixels has a color on the line R=B=G in RGB color space then determining that said image is a grey-scale anti-aliased image.
15. The method for detecting anti-aliased images of claim 10 , wherein; said color space is YUV space.
16. The method for detecting anti-aliased images of claim 10 , wherein; said color space is YCbCr space.