Programmable sample filtering for image rendering

• US 6,459,428 B1
• Filed: 10/03/2001
• Issued: 10/01/2002
• Est. Priority Date: 02/17/1998
• Status: Expired due to Term

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Deering, Michael F., Lavelle, Michael G., Oberoi, Ranjit S., Burk, Wayne Eric, Leung, Philip C., Tang, Yan Y.
• Primary Examiner(s): Zimmerman, Mark
• Assistant Examiner(s): Santiago, Enrique L.
• Application Number:
  US09/970,077
• Publication Number:
  US 20020101417A1
• Time in Patent Office: 363 Days
• Field of Search: 345/419, 345/420, 345/421, 345/422, 345/423, 345/424, 345/426, 345/427, 345/428, 345/581, 345/589, 345/611, 345/612, 345/613, 345/614, 345/615, 345/629, 345/473
• US Class Current: 345/419
• CPC Class Codes: G06T 11/001 Texturing; Colouring; Generation of texture or colour inpainting G06T5/77
• CPC Class Codes: G06T 15/503 Blending, e.g. for anti-aliasing
• CPC Class Codes: G06T 5/20 using local operators

Abstract

A graphics system configured to perform programmable filtering of samples to generate pixel values. The graphics system comprises a frame buffer, an accelerator unit and a video output processor. The accelerator unit receives graphics primitives, renders samples for the graphics primitives, and stores the rendered samples into a sample area of the frame buffer. The accelerator unit subsequently reads the samples from the sample area of the frame buffer, and filters the samples with a programmable filter having a programmable support region. The resulting pixel values are stored in a pixel area of the frame buffer. The video output processor reads the pixel values from the pixel area and converts the pixel values into a video signal which is provided to a video output port.

Claims

1. A graphics system comprising:

a frame buffer;

an accelerator unit configured to receive graphics primitives, render samples for the graphics primitives, and store the samples into a first area of the frame buffer;

wherein the accelerator unit is further configured to (a) read samples from the first area of the frame buffer, (b) filter the samples with a programmable filter to generate pixel values, and (c) store the pixel values in a second area of the frame buffer;

a video output processor configured to (d) read said pixel values from the second area of the frame buffer and (e) generate at least a portion of a video signal from said pixel values.

2. The graphics system of claim 1 further comprising a display device, wherein the video output processor is configured to provide said video signal to the display device.

3. The graphics system of claim 1 wherein the accelerator unit comprises a filter weight lookup table and a sample filter unit,

wherein the filter weight lookup table is implemented in RAM and stores filter weights defining the programmable filter,

wherein the sample filter unit is configured to read said filter weights from the filter weight lookup table and filter said samples with one or more of said filter weights to generate said pixel values.

4. The graphics system of claim 3, wherein said filter weights of said filter weight lookup table are modifiable by a host software program executing on a host computer.

5. The graphics system of claim 3, wherein the sample filter unit is configured to compute normalized square distances for said samples with respect to a current pixel position, to determine which of said samples reside interior to a filter support region corresponding to said programmable filter by comparing said normalized square distances to an upper limit value.

6. The graphics system of claim 5, wherein the sample filter unit is configured to compute the normalized square distance for a first of said samples by (i) computing a sum of squares of horizontal and vertical displacements of the first sample with respect to the current pixel position and (ii) multiplying said sum of squares by the reciprocal of the square of a radius of the filter support region.

7. The graphics system of claim 6, wherein the reciprocal of the square of the radius of the filter support region is stored in a programmable register.

8. The graphics system of claim 5, wherein the sample filter unit is configured to perform (b) by:

assigning a weight value of zero for each sample that resides exterior to the filter

support region;

assigning a weight value for each sample interior to the filter support region by accessing the filter weight lookup table with the corresponding normalized square distance;

multiplying sample attributes for each of said samples by the corresponding weight value to obtain weighted sample attributes;

accumulating one or more sums of weighted sample attributes;

accumulating a sum of said weight values of said samples;

multiplying each of said one or more sums of weighted sample attributes by a reciprocal of said sum of said weight values.

9. The graphics system of claim 5, wherein the sample filter unit is configured to perform (b) by:

assigning a weight value for each of said samples residing interior to the filter support region by accessing the filter weight lookup table with the corresponding normalized square distance;

multiplying sample attributes for each of said interior samples by the corresponding weight value to obtain weighted sample attributes;

accumulating one or more sums of the weighted sample attributes corresponding to the interior samples;

accumulating a sum of said weight values corresponding to said interior samples;

multiplying each of said one or more sums of weighted sample attributes by a reciprocal of said sum of said weight values.

10. The graphics system of claim 1, wherein said samples correspond to a set of bins which cover a filter support region centered at a current pixel position.

11. The graphics system claim 1, wherein said accelerator unit is programmable to set the programmable filter to a box filter having a rectangular filter support region, wherein said accelerator unit is configured to filter said samples by applying a constant weighting to a subset of said samples residing interior to the rectangular filter support region.

12. The graphics system of claim 11, wherein the rectangular filter support region is a square.

13. The graphics system of claim 1, wherein (c) and (d) are performed asynchronously and a different clock rates.

14. A method for generating displayable images, the method comprising:

(a) receiving a stream of graphics primitives;

(b) rendering samples for the graphics primitives;

(c) storing the samples into a first area of a frame buffer;

(d) reading samples from the first area of the frame buffer;

(e) filtering the samples with a programmable filter which covers a programmable support region to generate pixel values;

(f) storing the pixel values in a second area of the frame buffer;

(g) reading the pixel values from the second area of the frame buffer; and

(h) generating a portion of a video signal from the pixel values and providing said video signal to a display device, wherein (b) and (e) are performed by dedicated circuitry configured within a single integrated circuit.

15. The method of claim 14 wherein (e) comprises:

reading filter weights from a set of filter weights stored in a filter weight lookup table, wherein the set of filter weights define the programmable filter;

and filtering said samples with one or more of said filter weights to generate said pixel values.

16. The method of claim 15 further comprising a host computer writing said set of filter weights into said filter weight lookup table.

17. The method of claim 15, wherein (e) further comprises:

computing normalized square distances for said samples with respect to a current pixel position;

determining which of said samples reside interior to the programmable support region by comparing said normalized square distances to an upper limit value.

18. The method of claim 17, wherein said computing the normalized square distance for a first of said samples comprises:

(i) computing a sum of squares of horizontal and vertical displacements of the first sample with respect to the current pixel position and (ii) multiplying said sum of squares by the reciprocal of the square of a radius of the programmable support region.

19. The method of claim 17, wherein (e) comprises:

assigning a weight value of zero for each sample that resides exterior to the programmable support region;

assigning a weight value for each sample interior to the filter support region by reading the filter weight lookup table with the corresponding normalized square distance;

multiplying sample attributes for each of said samples by the corresponding weight value to obtain weighted sample attributes;

accumulating one or more sums of said weighted sample attributes corresponding to said samples;

accumulating a sum of said weight values of said samples; and

multiplying said one or more sums of weighted sample attributes by a reciprocal of said sum of said weight values.

20. The method of claim 17, wherein (e) comprises:

assigning a weight value for each of said interior samples by reading the filter weight lookup table with the corresponding normalized square distance;

multiplying sample attributes for each of said interior samples by the corresponding weight value to obtain weighted sample attributes;

accumulating one or more sums of said weighted sample attributes corresponding to said interior samples;

accumulating a sum of said weight values of said samples; and

multiplying said one or more sums of weighted sample attributes by a reciprocal of said sum of weight values.

21. The method claim 14 wherein (e) comprises:

setting the programmable support region to a rectangular region and setting the programmable filter to a constant filter, and

applying the constant filter to a subset of said sample residing interior to the rectangular region to generate said pixel values.

22. The method of claim 14 wherein (f) and (g) are performed asynchronously and at different clock rates.

23. A memory medium comprising computer readable instructions, wherein the instructions are executable to implement:

(a) receiving a stream of graphics primitives;

(b) rendering samples for the graphics primitives;

(c) storing the samples into a first area of a frame buffer;

(d) reading samples from the first area of the frame buffer;

(e) filtering the samples with a programmable filter which covers a programmable support region to generate pixel values;

(f) storing the pixel values in a second area of the frame buffer;

(g) reading the pixel values from the second area of the frame buffer; and

(h) generating a portion of a video signal from the pixel values and providing said video signal to a display device, wherein (b) and (e) are performed by dedicated circuitry configured within a single integrated circuit.