Methods and apparatus for improving quality of block-transform coded images
Block-based image processing methods and apparatus that provide a reduction in block-transform image coding artifacts are described. In various embodiments, the invention is directly incorporated into a decoding process. In such embodiments, transform coefficients are modified in simple but particularly effective ways that reduce or eliminate many of the artifacts that were caused by the quantization of the transform coefficients during encoding and/or by independent block processing during encoding. In other embodiments, the invention is used on an image that has already been decoded. In such embodiments image data values are directly modified in a block-based fashion or a forward block transform is applied and then the methods of the invention for processing transform coefficient blocks are used, followed by an inverse transform operation to generate pixel values from the resulting processed transform coefficient blocks.
1. A method of processing image data, the image data including blocks of transform coefficients corresponding to different portions of an image, each block of transform coefficients including at least one AC coefficient, each transform coefficient having a value, the method comprising:
examining the value of each one of a plurality of AC coefficients from one of said blocks to identify coefficients; and
filtering a set of AC coefficients including a first identified AC coefficient from said one of said blocks and a corresponding AC coefficient from at least one neighboring block to generate a filtered AC coefficient.
2. The method of claim 1 , further comprising:
processing the filtered AC coefficient to confine the AC coefficient to a quantization interval corresponding to the first AC coefficient.
3. The method of claim 2 , further comprising:
incorporating the processed filtered AC coefficient into a processed block of coefficients representing the same image portion as said one of said blocks.
4. The method of claim 3 , wherein said coefficients included in each block are quantized transform coefficients, each quantized transform coefficient having a value equal to the value of a quantization level.
5. The method of claim 3 , wherein each of the transform coefficients corresponds to a quantization interval, transform coefficients corresponding to the same quantization interval having the same value.
6. The method of claim 5 , further comprising:
retrieving information defining said quantization intervals from an encoded data stream which includes said blocks of transform coefficients.
7. The method of claim 5 , further comprising:
deriving said quantization intervals from the transform coefficients included in said blocks of transform coefficients.
8. The method of claim 3 , wherein examining each one of the plurality of AC coefficients includes comparing the examined AC coefficient to zero.
9. The method of claim 3 , wherein examined AC transform coefficients, which are not identified to be processed using AC transform coefficient filtering are represented in said processed block by coefficients generated from predictions made from transform coefficient values.
10. A set of computer executable instructions for controlling a processor to perform the steps recited in claim 3 embodied in a computer-readable storage medium.
11. The method of claim 3 , wherein the transform is a discrete cosine transform.
12. The method of claim 3 , wherein said filtering includes operating a finite impulse response filter of finite spatial extent to filter said set of AC coefficients.
13. The method of claim 2 , wherein said processing includes clamping the filtered AC coefficient to the nearest end value of said quantization interval when said filtered AC coefficient falls outside said quantization interval.
14. The method of claim 1 , wherein said plurality of AC coefficients from one of said blocks corresponds to particular pre-selected coefficient locations in said block.
15. The method of claim 14 , wherein the particular pre-selected coefficient locations in said block are determined as a function of the location of the portion of the image corresponding to said one of said blocks.
16. The method of claim 1 , wherein the step of examining the value of each one of a plurality of AC coefficients includes:
comparing the value of each one of the plurality of AC coefficients to a value; and
identifying AC coefficients as a function of the result of said comparison.
17. A system for processing image data, the image data including blocks of transform coefficients corresponding to different portions of an image, each block of transform coefficients including at least one AC coefficient, each transform coefficient having a value, the system comprising:
means for examining the value of each one of a plurality of AC coefficients from one of said blocks to identify coefficients;
means for filtering a set of AC coefficients including a first identified AC coefficient from said one of said blocks and a corresponding AC coefficient from at least one neighboring block to generate a filtered AC coefficient;
means for processing the filtered AC coefficient to confine the AC coefficient to a quantization interval corresponding to the first AC coefficient; and
means for incorporating the processed filtered AC coefficient into a processed block of coefficients representing the same image portion as said one of said blocks.
18. A method of processing image data, the image data including blocks of transform coefficients corresponding to different portions of an image, each block of transform coefficients including at least one AC coefficient, each transform coefficient having a value, the method comprising:
identifying coefficients as a function of the block in which the coefficient is included; and
filtering a set of AC coefficients including a first identified AC coefficient from one of said blocks and a corresponding AC coefficient from at least one neighboring block to generate a filtered AC coefficient.
19. The method of claim 18 , further comprising:
processing the filtered AC coefficient to confine the AC coefficient to a quantization interval corresponding to the first AC coefficient; and
incorporating the processed filtered AC coefficient into a processed block of coefficients representing the same image portion as said one of said blocks.
20. The method of claim 19 ,
wherein the step of identifying coefficients includes considering different coefficients in different blocks for identification purposes.
21. The method of claim 20 , wherein the step of considering different coefficients includes:
considering a first set of coefficients included in a first block; and
considering a second set of coefficients included in a second block, the second set of coefficients having at least one coefficient that is located at a different position within the block than the coefficients included in the first block.
22. The method of claim 20 , wherein said processing includes clamping the filtered AC coefficient to the nearest end value of said quantization interval when said filtered AC coefficient falls outside said quantization interval.
23. The method of claim 20 , wherein said coefficients included in each block are quantized transform coefficients, each quantized transform coefficient having a value equal to the value of a quantization level.
24. The method of claim 20 , wherein each of the transform coefficients corresponds to a quantization interval, transform coefficients corresponding to the same quantization interval having the same value.
25. The method of claim 24 , further comprising:
retrieving information defining said quantization intervals from an encoded data stream which includes said blocks of transform coefficients.
26. The method of claim 24 , further comprising:
deriving said quantization intervals from the transform coefficients included in said blocks of transform coefficients.
27. The method of claim 18 , wherein the transform is a discrete cosine transform.
28. The method of claim 18 , wherein said filtering includes operating a finite impulse response filter of finite spatial extent to filter said set of AC coefficients.
29. A set of computer executable instructions for controlling a processor to perform the steps recited in claim 18 embodied in a computer-readable storage medium.