Modified high dynamic range color decompression

• US 7,920,749 B1
• Filed: 09/17/2007
• Issued: 04/05/2011
• Est. Priority Date: 12/20/2006
• Status: Active Grant

• Current Assignee: NVIDIA Corporation
• Original Assignee: NVIDIA Corporation
• Inventors: Donovan, Walter E.
• Primary Examiner(s): Ge; Yuzhen
• Application Number:
  US11/856,550
• Time in Patent Office: 1,296 Days
• Field of Search: None
• US Class Current: 382/233
• CPC Class Codes: H04N 19/105 Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
• CPC Class Codes: H04N 19/184 the unit being bits, e.g. of the compressed video stream
• CPC Class Codes: H04N 19/186 the unit being a colour or a chrominance component
• CPC Class Codes: H04N 19/593 involving spatial prediction techniques

Abstract

Systems and methods for representing high dynamic range data in compressed formats with a fixed size block allow high dynamic range data to be stored in less memory. The compressed formats use 8 bits per pixel. A first compressed format includes two endpoint values and compressed indices for the pixels in the block. A second compressed format includes four endpoint values, a partition index that specifies a mask for each pair of the four endpoint values, and compressed indices for the pixels in the block. The two formats may be used for various blocks within a single compressed image and mode bits are included to distinguish between the two formats. Furthermore, each endpoint value may be encoded using an endpoint compression mode that is also specified by the mode bits. Compressed high dynamic range values represented in either format may be efficiently decompressed in hardware.

Claims

1. A method of decompressing high dynamic range data represented in a compressed format, comprising:

determining an endpoint compression mode that is specified by the compressed format;

extracting a first decompressed endpoint from a compressed endpoint field of the compressed format;

extracting one or more remaining compressed endpoints from the compressed endpoint field of the compressed format;

computing, using a processor, a second decompressed endpoint when the endpoint compression mode indicates that the one or more remaining compressed endpoints are encoded as differences;

computing a third decompressed endpoint and a fourth decompressed endpoint when the endpoint compression mode indicates that dual endpoint pair compression is used to encode the high dynamic range data;

extracting compressed indices from the compressed format;

decompressing the compressed indices to produce decompressed indices that include an index for each pixel in a fixed sized block; and

computing a decompressed high dynamic range value for each channel of each of the pixels in the fixed size block using the first decompressed endpoint and the second decompressed endpoint or the first decompressed endpoint, the second decompressed endpoint, the third decompressed endpoint, and the fourth decompressed endpoint and the decompressed indices.

2. The method of claim 1, further comprising:

obtaining the second decompressed endpoint from the one or more remaining compressed endpoints when the endpoint compression mode indicates that the one or more remaining compressed endpoints are not encoded as differences; and

obtaining the third decompressed endpoint and the fourth decompressed endpoint from the one or more remaining compressed endpoints when the endpoint compression mode indicates that dual endpoint pair compression is used to encode the high dynamic range data.

3. The method of claim 1, wherein the computing of the second decompressed endpoint comprises:

sign extending the one or more remaining compressed endpoints to produce a sign extended second compressed endpoint; and

summing the sign extended second compressed endpoint with the first decompressed endpoint to produce the second decompressed endpoint.

4. The method of claim 3, wherein the computing of the third decompressed endpoint and the fourth decompressed endpoint comprises:

sign extending the one or more remaining compressed endpoints to produce a sign extended third compressed endpoint and a sign extended fourth compressed endpoint;

summing the sign extended third compressed endpoint with the first decompressed endpoint to produce the third decompressed endpoint; and

summing the sign extended fourth compressed endpoint with the second decompressed endpoint to produce the fourth decompressed endpoint.

5. The method of claim 1, further comprising the step of determining a partition mask that is specified by the compressed format and includes a bit for each of the pixels in the fixed size block to indicate a portion of the pixels in the fixed size block that are compressed using the first decompressed endpoint and the second decompressed endpoint and a remaining portion of the pixels in the fixed size block that are compressed using the third decompressed endpoint and the fourth decompressed endpoint.

6. The method of claim 5, wherein the step of computing a decompressed high dynamic range value for each channel of each pixel in the fixed size block comprises:

generating a first palette including a plurality of decompressed high dynamic range values by interpolating between the first decompressed endpoint and the second decompressed endpoint using a set of weights, wherein each weight corresponds to a possible index value;

generating a second palette including a plurality of decompressed high dynamic range values by interpolating between the third decompressed endpoint and the fourth decompressed endpoint using the set of weights; and

selecting one of the plurality of decompressed high dynamic range values in the first palette or the second palette for each channel of each pixel in the fixed size block using a decompressed index for the pixel and the partition mask.

7. The method of claim 1, wherein the step of decompressing the compressed indices comprises determining a bit of one of the compressed indices based on a partition index that is specified by the compressed format.

8. The method of claim 1, wherein the step of computing a decompressed high dynamic range value for each channel of each pixel in the fixed size block comprises:

generating a palette including a plurality of decompressed high dynamic range values by interpolating between the first decompressed endpoint and the second decompressed endpoint using a set of weights, wherein each weight corresponds to a possible index value; and

selecting one of the plurality of decompressed high dynamic range values in the palette for each channel of each pixel in the fixed size block using a decompressed index for the pixel.

9. The method of claim 1, wherein the step of decompressing the compressed indices comprises appending a zero in the most significant bit position of one of the compressed indices.

10. The method of claim 1, further comprising the step of unquantizing channels of the first decompressed endpoint and channels of the second decompressed endpoint to produce 16 bit floating point or integer values that are signed or unsigned.

11. A system for decompressing high dynamic range data represented in a compressed format, comprising:

an endpoint computation unit, comprising a processor and a memory, configured to

determine an endpoint compression mode that is specified by the compressed format;

extract a first decompressed endpoint from a compressed endpoint field of the compressed format;

extract one or more remaining compressed endpoints from the compressed endpoint field of the compressed format;

compute a second decompressed endpoint when the endpoint compression mode indicates that the one or more remaining compressed endpoints are encoded as differences; and

compute a third decompressed endpoint and a fourth decompressed endpoint when the endpoint compression mode indicates that dual endpoint pair compression is used to encode the high dynamic range data;

an index computation unit configured to

extract compressed indices from the compressed format; and

decompress the compressed indices to produce decompressed indices that include an index for each pixel in a fixed sized block; and

a pixel value computation unit coupled to the endpoint computation unit and the index computation unit, and configured to compute a decompressed high dynamic range value for each channel of each of the pixels in a fixed size block using the first decompressed endpoint and the second decompressed endpoint or the first decompressed endpoint, the second decompressed endpoint, the third decompressed endpoint, and the fourth decompressed endpoint and the decompressed indices.

12. The system of claim 11, wherein the endpoint computation unit is further configured to unquantize channels of the first decompressed endpoint and channels of the second decompressed endpoint to produce 16 bit floating point or integer values that are signed or unsigned.

13. The system of claim 11, further comprising a partition mask table that is coupled to the index computation unit and configured to store partition masks corresponding to a partition index that is specified by the compressed format, wherein each partition mask includes a bit for each pixel in the fixed size block to indicate a portion of the pixels in the fixed size block that are compressed using the first decompressed endpoint and the second decompressed endpoint and a remaining portion of the pixels in the fixed size block that are compressed using the third decompressed endpoint and the fourth decompressed endpoint.

14. The system of claim 13, wherein the index computation unit is further configured to determine a bit of one of the compressed indices based on the partition index.

15. The system of claim 13, further comprising a palette generation unit configured to

generate a first palette including a plurality of decompressed high dynamic range values by interpolating between the first decompressed endpoint and the second decompressed endpoint using a set of weights, wherein each weight corresponds to a possible index value;

generate a second palette including a plurality of decompressed high dynamic range values by interpolating between the third decompressed endpoint and the fourth decompressed endpoint using the set of weights; and

select one of the plurality of decompressed high dynamic range values in the first palette or the second palette for each channel of each pixel in the fixed size block using a decompressed index for the pixel and the partition mask.

16. The system of claim 13, further comprising a palette generation unit configured to

generate a palette including a plurality of decompressed high dynamic range values by interpolating between the first decompressed endpoint and the second decompressed endpoint using a set of weights, wherein each weight corresponds to a possible index value;

and

select one of the plurality of decompressed high dynamic range values in the palette for each channel of each pixel in the fixed size block using a decompressed index for the pixel.

17. The system of claim 11, further comprising a texture filter unit that is coupled to the pixel value computation unit and configured to interpret the decompressed high dynamic range values for each channel of each pixel as values represented in a floating point format or a fixed point format.