Strategies for Performing Scaling Operations on Image Information

• US 20060153476A1
• Filed: 03/14/2006
• Published: 07/13/2006
• Est. Priority Date: 08/01/2003
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Evans, Glenn F., Spears, Stacey L., Munsil, Donald J.
• Granted Patent: US 7,400,762 B2
• Time in Patent Office: Days
• US Class Current: 382/299
• CPC Class Codes: H04N 1/4053 with threshold modulated relative to input image data or vice versa

Abstract

Strategies are set forth for performing a scaling operation on image information using a filter. According to one exemplary implementation, the strategies involve: computing, based on scaling to be performed, a number of filter kernels required to implement the filter and a number of taps required for each filter kernel; pre-computing and storing the number of filter kernels; and applying the filter kernels to a row or column of the image information by repeatedly sequencing through the filter kernels if there are more pixel elements in the row or column than there are pre-calculated filter kernels. The scaling operation can employ a Catmull-Rom filter. Further, the scaling operations can be implemented by a pixel shader by assigning weighting information and delta-shifted versions of the image information to respective texture units of the pixel shader.

Claims

1. A method for performing a scaling operation on image information using a filter, comprising:

computing, based on scaling to be performed, a number of filter kernels required to implement the filter;

pre-computing and storing the number of filter kernels; and

applying the pre-computed filter kernels to a subset of the image information.

2. The method of claim 1, wherein the scaling operation employs a Catmull-Rom filter.

3. The method of claim 1, wherein the computing also involves determining how many taps each kernel should have.

4. The method of claim 1, wherein the subset of the image information comprises a row of the image information.

5. The method of claim 1, wherein the subset of the image information comprises a column of the image information.

6. The method of claim 1, wherein the applying comprising repeatedly sequencing through the filter kernels if there are more pixel elements in the subset of image information than there are pre-computed filter kernels.

7. The method of claim 1, further comprising constraining the scaling operation to scaling options that do not require more than a prescribed number of filter kernels.

8. The method of claim 1, further comprising applying the pre-computed filter kernels that were computed for the subset of the image information to other subsets of the image information.

9. The method of claim 1, wherein the scaling operation is performed using a graphical processing unit that includes a pixel shader.

10. The method of claim 9, further comprising performing the scaling operation using the pixel shader by:

assigning weighting information to at least one texture unit;

assigning delta-shifted versions of the image information to plural other texture units; and

performing an operation using the pixel shader based on information received from the texture units, to provide an output result.

11. One or more computer readable media storing machine-readable instructions configured to implement the method of claim 1.

12. An apparatus for performing a scaling operation on image information using a filter, comprising:

a dynamic kernel pre-calculation module configured to;

compute, based on scaling to be performed, a number of filter kernels required to implement the filter; and

pre-compute and store the number of filter kernels;

a kernel storage module configured to store the pre-computed filter kernels; and

a kernel application module configured to apply the pre-computed filter kernels to a subset of the image information, wherein the applying comprising repeatedly sequencing through the filter kernels if there are more pixel elements in the subset of image information than there are pre-computed filter kernels.

13. The apparatus of claim 12, wherein the scaling operation employs a Catmull-Rom filter.

14. The apparatus of claim 12, further including a graphical processing unit that includes a pixel shader for performing the scaling operation.

15. The apparatus of claim 12, wherein the pixel shader includes plural texture units, and wherein the pixel shader is configured to perform the scaling operation by:

assigning weighting information to at least one texture unit;

assigning delta-shifted versions of the image information to plural other texture units; and

performing an operation using the pixel shader based on information received from the texture units, to provide an output result.

16. A method for performing at least one processing task in an image processing pipeline using a pixel shader of a graphical processing unit, comprising:

assigning weighting information to at least one texture unit of the pixel shader;

assigning delta-shifted versions of input image information to plural other texture units of the pixel shader; and

performing an operation using the pixel shader based on the information from the texture units, to thereby provide an output result.

17. The method of claim 16, further comprising performing plural operations using the pixel shader to generate a final output result on the basis of plural partial output results, wherein the performing of plural operations is applied to a case in which the scaling operation uses a filter having a number of taps which exceeds a number of available texture units.

18. The method of claim 16, further comprising applying a hide pixel mode to automatically hide edge conditions in image information that should not be displayed.

19. The method of claim 16, farther comprising applying a mirror mode to reflect image information either horizontally or vertically when referencing image information outside boundaries of the image information.

20. One or more computer readable media storing machine-readable instructions configured to implement the method of claim 16.