I/WE CLAIM:
1. A method comprising:
sampling multiple locations within each pixel of an image frame to generate a plurality of image slices, each image slice comprising a different set of samples of the pixels of the image frame;
performing anti-aliasing of the image frame using the image slices by performing the operations of:
determining whether one or more different pixel block values are different in one or more different image slices;
performing edge detection and shape detection using pixel data from a pixel block in a single image slice if the pixel block does not include different pixel values in different image slices; and
performing edge detection and simple shape detection using the pixel block in multiple image slices if the pixel block includes different pixel values in different image slices.
2. The method of claim 1 wherein the operation of sampling multiple locations within each pixel comprises an operation in a Multisample Anti-Aliasing (MSAA) sequence of operations.
3. The method of claim 1 further comprising:
resolving the plurality of image slices into a single anti-aliased image frame.
4. The method of claim 3 further comprising:
blending pixels from the image slices to generate the single anti-aliased image frame.
5. The method of claim 1 wherein performing edge detection comprises:

identifying an edge when a difference of neighboring pixel values compared to pixel values of the edge is above a preset threshold.
6. The method of claim 5 further comprising:
pruning one or more non-dominant edges by comparing color delta values of the non-dominant edges to that of a plurality of closest neighboring edges.
7. The method of claim 5 wherein simple shape detection comprises detecting two edge shapes, three edge shapes, four edge shapes and Z-shapes.
8. The method of claim 7 wherein one or more of the Z-shapes are non-symmetric.
9. The method of claim 1 wherein performing edge detection comprises:
suppressing horizontal and vertical edge strength values are using a
plurality of connecting edge strength values of an opposite orientation.
10. The method of claim 1 wherein performing edge detection comprises
performing local contrast adaptation.
11. The method of claim 10 wherein local contrast adaptation comprises:
reducing each horizontal edge strength by a sum of a plurality of
connecting vertical edge strengths multiplied by a predetermined value; and
reducing each vertical edge strength by a sum of a plurality of connecting horizontal edge strengths multiplied by a predetermined value.
12. The method of claim 1 further comprising:
blending anti-aliased color values from the plurality of slices to generate a single anti-aliased image frame.
13. A graphics processing apparatus comprising:

a sampler to sample multiple locations within each pixel of an image frame to generate a plurality of image slices, each image slice comprising a different set of samples of the pixels of the image frame;
anti-aliasing logic to perform anti-aliasing of the image frame using the image slices by performing the operations of:
determining whether one or more different pixel block values are different in one or more different image slices;
performing edge detection and shape detection using pixel data from a pixel block in a single image slice if the pixel block does not include different pixel values in different image slices; and
performing edge detection and simple shape detection using the pixel block in multiple image slices if the pixel block includes different pixel values in different image slices.
14. The graphics processing apparatus of claim 13 wherein the operation of sampling multiple locations within each pixel comprises an operation in a Multisample Anti-Aliasing (MSAA) sequence of operations.
15. The graphics processing apparatus of claim 13 wherein the anti-aliasing circuitry is to resolve the plurality of image slices into a single anti-aliased image frame.
16. The graphics processing apparatus of claim 15 further comprising:
blending logic to blend pixels from the image slices to generate the single
anti-aliased image frame.
17. The graphics processing apparatus of claim 13 wherein performing
edge detection comprises identifying an edge when a difference of neighboring
pixel values compared to pixel values of the edge is above a preset threshold.

18. The graphics processing apparatus of claim 17 further comprising:
pruning logic to prune one or more non-dominant edges by comparing
color delta values of the non-dominant edges to that of a plurality of closest neighboring edges.
19. The graphics processing apparatus of claim 17 wherein simple shape detection comprises detecting two edge shapes, three edge shapes, four edge shapes and Z-shapes.
20. The graphics processing apparatus of claim 19 wherein one or more of the Z-shapes are non-symmetric.
21. The graphics processing apparatus of claim 13 wherein performing edge detection comprises:
suppressing horizontal and vertical edge strength values are using a plurality of connecting edge strength values of an opposite orientation.
22. The graphics processing apparatus of claim 13 wherein performing edge detection comprises performing local contrast adaptation.
23. The graphics processing apparatus of claim 22 wherein local contrast adaptation comprises:
reducing each horizontal edge strength by a sum of a plurality of connecting vertical edge strengths multiplied by a predetermined value; and
reducing each vertical edge strength by a sum of a plurality of connecting horizontal edge strengths multiplied by a predetermined value.
24. The graphics processing apparatus of claim 13 further comprising:

blending logic to blend anti-aliased color values from the plurality of slices to generate a single anti-aliased image frame.
25. A machine-readable medium having program code stored thereon
which, when executed by a machine, causes the machine to perform the operations
of:
sampling multiple locations within each pixel of an image frame to generate a plurality of image slices, each image slice comprising a different set of samples of the pixels of the image frame;
performing anti-aliasing of the image frame using the image slices by performing the operations of:
determining whether one or more different pixel block values are different in one or more different image slices;
performing edge detection and shape detection using pixel data from a pixel block in a single image slice if the pixel block does not include different pixel values in different image slices; and
performing edge detection and simple shape detection using the pixel block in multiple image slices if the pixel block includes different pixel values in different image slices.
26. The machine-readable medium of claim 25 wherein the operation of sampling multiple locations within each pixel comprises an operation in a Multisample Anti-Aliasing (MSAA) sequence of operations.
27. The machine-readable medium of claim 26 further comprising program code to cause the machine to perform the operation of:
resolving the plurality of image slices into a single anti-aliased image frame.

28. The machine-readable medium of claim 27 further comprising program
code to cause the machine to perform the operation of:
blending pixels from the image slices to generate the single anti-aliased image frame.
29. The machine-readable medium of claim 25 wherein performing edge
detection comprises:
identifying an edge when a difference of neighboring pixel values compared to pixel values of the edge is above a preset threshold.
30. The machine-readable medium of claim 29 further comprising program
code to cause the machine to perform the operation of:
pruning one or more non-dominant edges by comparing color delta values of the non-dominant edges to that of a plurality of closest neighboring edges.
31. The machine-readable medium of claim 29 wherein simple shape detection comprises detecting two edge shapes, three edge shapes, four edge shapes and Z-shapes.
32. The machine-readable medium of claim 31 wherein one or more of the Z-shapes are non-symmetric.
33. The machine-readable medium of claim 25 wherein performing edge detection comprises:
suppressing horizontal and vertical edge strength values are using a plurality of connecting edge strength values of an opposite orientation.
34. The machine-readable medium of claim 25 wherein performing edge
detection comprises performing local contrast adaptation.

35. The machine-readable medium of claim 34 wherein local contrast
adaptation comprises:
reducing each horizontal edge strength by a sum of a plurality of connecting vertical edge strengths multiplied by a predetermined value; and
reducing each vertical edge strength by a sum of a plurality of connecting horizontal edge strengths multiplied by a predetermined value.
36. The machine-readable medium of claim 25 further comprising program
code to cause the machine to perform the operation of:
blending anti-aliased color values from the plurality of slices to generate a single anti-aliased image frame.