Method and apparatus for controlling amount of quantization processing in an encoder

• US 7,042,941 B1
• Filed: 07/17/2001
• Issued: 05/09/2006
• Est. Priority Date: 07/17/2001
• Status: Active Grant

• Current Assignee: VIXS Systems Incorporated (Pixelworks Incorporated), VIXS Incorporated
• Original Assignee: VIXS Incorporated
• Inventors: Chan, Jason, Laksono, Indra
• Primary Examiner(s): Philippe, Gims
• Application Number:
  US09/907,033
• Time in Patent Office: 1,757 Days
• Field of Search: 375/240.03, 375/240.05, 375/240.16, 375/240.11, 375/240.22, 348/395.1, 348/404.1, 382/251, 382/253
• US Class Current: 375/240.03
• CPC Class Codes: H04N 19/124 Quantisation
• CPC Class Codes: H04N 19/152 by measuring the fullness of the transmission buffer
• CPC Class Codes: H04N 19/154 Measured or subjectively estimated visual quality after decoding, e.g. measurement of distortion use of rate-distortion criteria H04N19/147
• CPC Class Codes: H04N 19/159 Prediction type, e.g. intra-frame, inter-frame or bidirectional frame prediction
• CPC Class Codes: H04N 19/172 the region being a picture, frame or field
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A method and apparatus for controlling the amount of quantization processing used within an encoder that compresses video and/or audio data include processing that begins by receiving discrete cosine transform data of a block of a frame of data. The process then proceeds by obtaining a quantization table related to the frame of data. The processing then continues by obtaining a quantization scaling factor related to the frame. The processing then continues by determining whether quantization processing limits have been exceeded for quantization of preceding blocks of the frame of data. If the number of bits processed for the preceding blocks exceeds a certain desired level, the processing increases the quantization scaling factor. The processing then continues by generating quantization data from the discrete cosine transform data, the quantization table and the increased quantization scaling factor.

Claims

1. A method for controlling amount of quantization processing, the method comprises:

receiving discrete cosine transform data of a block of a frame of data, wherein the frame of data is divided into a plurality of regions, wherein each of the plurality of regions includes a plurality of blocks, and wherein the block is within one of the plurality of regions;

obtaining a quantization table relating to the frame of data;

obtaining an allocated number of bits for quantizing data for each of the plurality of regions;

obtaining a plurality of quantization scaling factors relating to the frame of data;

generating a plurality of quantization matrixes based on the plurality of quantization scaling factors, the quantization table, and the discrete cosine transform data;

for the block of the frame of data, determining whether an actual number of bits to quantize data of a preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data for the preceding one of the plurality of regions; and

selecting one of the plurality of quantization matrixes as quantization data based on a decision input and the determining whether the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data of the preceding one of the plurality of regions.

2. The method of claim 1 further comprises:

for other blocks the one of the plurality of regions;

generating another plurality of quantization matrixes based on the plurality of quantization scaling factors, the quantization table, and the discrete cosine transform data of the another block of the other blocks; and

selecting one of the another plurality of quantization matrixes as quantization data for the another block based on the decision input and the determining whether the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data of the preceding one of the plurality of regions.

3. The method of claim 1, wherein the selecting one of the plurality of quantization matrixes further comprises:

selecting an initial one of the plurality of quantization matrixes based on the decision input; and

when the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeds the allocated number of bits for quantizing data of the preceding one of the plurality of regions, selecting the one of the plurality of quantization matrixes based the initial one of the plurality of quantization matrixes.

4. A method for controlling amount of quantization processing, the method comprises:

partitioning a frame of MPEG encoded data into a plurality of regions, wherein each of the plurality of regions includes a plurality of blocks;

assigning a desired bit processing value to each of the plurality of regions;

monitoring quantization of the blocks of one of the plurality of regions to obtain bits used to quantize the blocks of the one of the plurality of regions, wherein the quantization includes selecting one of a plurality of quantization matrixes based on a decision input; and

when the bits used to quantize the blocks of the one of the plurality of regions exceeds the desired bit processing value for the region, adjusting the decision input for quantizing data of at least one subsequent region of the plurality of regions.

5. The method of claim 4, wherein the partitioning of the frame of MPEG encoded data further comprises at least one of:

partitioning the frame of MEPG encoded data on a per macroblock basis, such that each macroblock is within a unique one of the plurality of regions, wherein a macroblock includes at least a subset of the blocks of the unique one of the plurality of regions; and

partitioning the frame of MPEG encoded data into N number of the plurality of regions, where each region of the plurality of regions includes at least two macroblocks of the frame of the MPEG encoded data.

6. The method of claim 4, wherein the assigning the desired bit processing value further comprises:

dividing a total number of desired bits for quantizing the frame by number of regions in the plurality of regions to obtain the desired bit processing value.

7. The method of claim 4, wherein the adjusting the decision input further comprises:

determining amount in which the bits used to quantize the blocks of the one of the plurality of regions exceeds the desired bit processing value for the region to produce excess amount;

determining whether the excess amount exceeds a threshold;

when the excess amount does not exceed the threshold, utilizing a first increased quantization factor; and

when the excess amount exceeds the threshold, utilizing a second increased quantization factor, wherein the second increased quantization factor is greater than the first quantization factor, wherein the first and second quantization factors correspond to a level of insubstantial data.

8. A method for encoding data, the method comprises:

performing motion estimation upon a block of data based on motion vectors to produce motion estimated data;

performing a discrete cosine transform function upon the motion estimated data to produce discrete cosine transform data;

obtaining a quantization table;

obtaining an allocated number of bits for quantizing data for each of the plurality of regions;

obtaining a plurality of quantization scaling factors relating to the frame of data;

generating a plurality of quantization matrixes based on the plurality of quantization scaling factors, the quantization table, and the discrete cosine transform data;

for the block of the frame of data, determining whether an actual number of bits to quantize data of a preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data for the preceding one of the plurality of regions;

selecting one of the plurality of quantization matrixes as quantization data based on a decision input and the determining whether the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data of the preceding one of the plurality of regions;

performing a zigzag function upon the quantization quantized data to produce zigzag data; and

run level encoding the zigzag data to produce encoded data.

9. The method of claim 8 further comprises:

for other blocks of the one of the plurality of regions;

generating another plurality of quantization matrixes based on the plurality of quantization scaling factors, the quantization table, and the discrete cosine transform data of the another block of the other blocks; and

selecting one of the another plurality of quantization matrixes as quantization data for the another block based on the decision input and the determining whether the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data of the preceding one of the plurality of regions.

10. The method of claim 8, wherein the selecting one of the plurality of quantization matrixes further comprises:

selecting an initial one of the plurality of quantization matrixes based on the decision input; and

when the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeds the allocated number of bits for quantizing data of the preceding one of the plurality of regions, selecting the one of the plurality of quantization matrixes based the initial one of the plurality of quantization matrixes.

11. An apparatus for controlling amount of quantization processing, the apparatus comprises:

processing module; and

memory operably coupled to the processing module, wherein the memory includes operational instructions that cause the processing module to;

receive discrete cosine transform data of a block of a frame of data, wherein the frame of data is divided into a plurality of regions, wherein each of the plurality of regions includes a plurality of blocks, and wherein the block is within one of the plurality of regions;

obtain a quantization table relating to the frame of data;

obtain an allocated number of bits for quantizing data for each of the plurality of regions;

obtain a plurality of quantization scaling factors relating to the frame of data;

generate a plurality of quantization matrixes based on the plurality of quantization scaling factors, the quantization table, and the discrete cosine transform data;

for the block of the frame of data, determine whether an actual number of bits to quantize data of a preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data for the preceding one of the plurality of regions; and

select one of the plurality of quantization matrixes as quantization data based on a decision input and the determining whether the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data of the preceding one of the plurality of regions.

12. The apparatus of claim 11, wherein the memory further comprises operational instructions that cause the processing module to:

for other blocks the one of the plurality of regions;

generate another plurality of quantization matrixes based on the plurality of quantization scaling factors, the quantization table, and the discrete cosine transform data of the another block of the other blocks; and

select one of the another plurality of quantization matrixes as quantization data for the another block based on the decision input and the determining whether the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data of the preceding one of the plurality of regions.

13. The apparatus of claim 11, wherein the memory further comprises operational instructions that cause the processing module to select one of the plurality of quantization matrixes factor by:

selecting an initial one of the plurality of quantization matrixes based on the decision input; and

when the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeds the allocated-number of bits for quantizing data of the preceding one of the plurality of regions, selecting the one of the plurality of quantization matrixes based the initial one of the plurality of quantization matrixes.

14. An apparatus for controlling amount of quantization processing, the apparatus comprises:

processing module; and

memory operably coupled to the processing module, wherein the memory includes operational instructions that cause the processing module to;

partition a frame of MPEG encoded data into a plurality of regions, wherein each of the plurality of regions includes a plurality of blocks;

assign a desired bit processing value to each of the plurality of regions;

monitor quantization of the blocks of one of the plurality of regions to obtain bits used to quantize the blocks of the one of the plurality of regions, wherein the quantization includes selecting one of a plurality of quantization matrixes based on a decision input; and

when the bits used to quantize the blocks of the one of the plurality of regions exceeds the desired bit processing value for the region, adjust the decision input for quantizing data of at least one subsequent region of the plurality of regions.

15. The apparatus of claim 14, wherein the memory further comprises operational instructions that cause the processing module to partition of the frame of MPEG encoded data by at least one of:

partitioning the frame of MEPG encoded data on a per macroblock basis, such that each macroblock is within a unique one of the plurality of regions, wherein a macroblock includes at least a subset of the blocks of the unique one of the plurality of regions; and

partitioning the frame of MPEG encoded data into N number of the plurality of regions, where each region of the plurality of regions includes at least two macroblocks of the frame of the MPEG encoded data.

16. The apparatus of claim 14, wherein the memory further comprises operational instructions that cause the processing module to assign the desired bit processing value by:

dividing a total number of desired bits for quantizing the frame by number of regions in the plurality of regions to obtain the desired bit processing value.

17. The apparatus of claim 14, wherein the memory further comprises operational instructions that cause the processing module to adjust the decision input by:

determining amount in which the bits used to quantize the blocks of the one of the plurality of regions exceeds the desired bit processing value for the region to produce excess amount;

determining whether the excess amount exceeds a threshold;

when the excess amount does not exceed the threshold, utilizing a first increased quantization factor; and

when the excess amount exceeds the threshold, utilizing a second increased quantization factor, wherein the second increased quantization factor is greater than the first quantization factor, wherein the first and second quantization factors correspond to a level of insubstantial data.

18. An apparatus for encoding data, the apparatus comprises:

processing module; and

memory operably coupled to the processing module, wherein the memory includes operational instructions that cause the processing module to;

performing motion estimation upon a block of data based on motion vectors to produce motion estimated data;

perform a discrete cosine transform function upon the motion estimated data to produce discrete cosine transform data;

obtain a quantization table;

obtain a quantization scaling factor;

obtain an allocated number of bits for quantizing data for each of the plurality of regions;

obtain a plurality of quantization scaling factors relating to the frame of data;

generate a plurality of quantization matrixes based on the plurality of quantization scaling factors, the quantization table, and the discrete cosine transform data;

for the block of the frame of data, determine whether an actual number of bits to quantize data of a preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data for the preceding one of the plurality of regions;

select one of the plurality of quantization matrixes as quantization data based on a decision input and the determining whether the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data of the preceding one of the plurality of regions;

perform a zigzag function upon the quantization data to produce zigzag data; and

run level encoding the zigzag data to produce encoded data.

19. The apparatus of claim 18, wherein the memory further comprises operational instructions that cause the processing module to:

for other blocks of the one of the plurality of regions;

generate another plurality of quantization matrixes based on the plurality of quantization scaling factors, the quantization table, and the discrete cosine transform data of the another block of the other blocks; and

select one of the another plurality of quantization matrixes as quantization data for the another block based on the decision input and the determining whether the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeded the allocated number of bits for quantizing data of the preceding one of the plurality of regions.

20. The apparatus of claim 18, wherein the memory further comprises operational instructions that cause the processing module to select one of the plurality of quantization matrixes by:

selecting an initial one of the plurality of quantization matrixes based on the decision input; and

when the actual number of bits to quantize the data of the preceding one of the plurality of regions exceeds the allocated number of bits for quantizing data of the preceding one of the plurality of regions, selecting the one of the plurality of quantization matrixes based the initial one of the plurality of quantization matrixes.