Adaptive bit rate control for rate reduction of MPEG coded video

• US 6,937,770 B1
• Filed: 12/28/2000
• Issued: 08/30/2005
• Est. Priority Date: 12/28/2000
• Status: Expired due to Term

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Faibish, Sorin, Duso, Wayne W., Oguz, Seyfullah H.
• Primary Examiner(s): Wu, Jingge
• Application Number:
  US09/750,565
• Time in Patent Office: 1,706 Days
• Field of Search: 382232-239, 37524001-24026
• US Class Current: 382/235
• CPC Class Codes: H04N 19/132 Sampling, masking or truncation of coding units, e.g. adaptive resampling, frame skipping, frame interpolation or high-frequency transform coefficient masking
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/18 the unit being a set of transform coefficients
• CPC Class Codes: H04N 19/40 using video transcoding, i.e. partial or full decoding of a coded input stream followed by re-encoding of the decoded output stream
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

Original-quality MPEG coded video is processed to produce reduced-quality MPEG coded video at a reduced bit rate. The processing is based on a scale factor between average frame size of the original-quality MPEG coded video and a desired average frame size of the reduced-quality MPEG coded video. For each Discrete Cosine Transform (DCT) block of each frame, the processing calculates a size of the block of the reduced frame by scaling the original block size by the scale factor, and removes a sufficient number of bits from the original block to obtain substantially the calculated size. In addition, the processing accumulates excess bits when the block size reduction eliminates more bits from a block than are necessary for the desired reduction of the size of the block, and any excess bits are used for processing a number of following blocks.

Claims

1. A method of producing reduced-quality MPEG coded video from original-quality MPEG coded video, the original-quality MPEG coded video including a set of non-zero AC discrete cosine transform (DCT) coefficients for 8×

8 blocks in frames of the original-quality MPEG coded video, the reduced-quality MPEG coded video also having frames of 8×

8 blocks, each frame in the reduced-quality MPEG coded video having a corresponding frame in the original-quality MPEG coded video, and each 8×

8 block in each frame of the reduced-quality MPEG coded video having a corresponding block in a corresponding frame in the original-quality MPEG coded video, said method comprising the steps of;

selecting non-zero AC DCT coefficients from corresponding blocks in the original-quality MPEG coded video to be included in the blocks of the reduced-quality MPEG coded video so that each 8×

8 block in each frame of the reduced-quality MPEG coded video has a number of bits encoding non-zero AC DCT coefficients that is generally proportional to the number of bits encoding non-zero AC DCT coefficients for the corresponding 8×

8 block in the corresponding frame of original-quality MPEG coded video;

which includes computing a number of bits available for encoding the non-zero AC DCT coefficients in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video, the number of bits available for encoding being computed by scaling the number of bits encoding non-zero AC DCT coefficients in the corresponding block of the corresponding frame of the original-quality MPEG coded video by a scale factor for said each frame, the scale factor being selected to obtain a desired reduced bit rate in the MPEG coded video for said each frame; and

which includes computing a moving average of frame size of the frames in the original-quality MPEG coded video, and computing the scale factor for said each frame from the moving average of frame size and a desired frame size for the reduced-quality MPEG coded video.

2. A method of producing reduced-quality MPEG coded video from original-quality MPEG coded video, the original-quality MPEG coded video including a set of non-zero AC discrete cosine transform (DCT) coefficients for 8×

8 blocks in frames of the original-quality MPEG coded video, the reduced-quality MPEG coded video also having frames of 8×

8 blocks, each frame in the reduced-quality MPEG coded video having a corresponding frame in the original-quality MPEG coded video, and each 8×

8 block in each frame of the reduced-Quality MPEG coded video having a corresponding block in a corresponding frame in the original-quality MPEG coded video, said method comprising the steps of;

selecting non-zero AC DCT coefficients from corresponding blocks in the original-quality MPEG coded video to be included in the blocks of the reduced-quality MPEG coded video so that each 8×

8 block in each frame of the reduced-quality MPEG coded video has a number of bits encoding non-zero AC DCT coefficients that is generally proportional to the number of bits encoding non-zero AC DCT coefficients for the corresponding 8×

8 block in the corresponding frame of original-quality MPEG coded video;

which includes computing a number of bits available for encoding the non-zero AC DCT coefficients in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video, the number of bits available for encoding being computed by scaling the number of bits encoding non-zero AC DCT coefficients in the corresponding block of the corresponding frame of the original-quality MPEG coded video by a scale factor for said each frame, the scale factor being selected to obtain a desired reduced bit rate in the MPEG coded video for said each frame; and

which includes computing a difference between the number of bits available for encoding the non-zero AC DCT coefficients in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video and the number of bits used for encoding the non-zero AC DCT coefficients retained in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video, and making available for encoding non-zero AC DCT coefficients of following blocks said difference between the number of bits.

3. The method as claimed in claim 2, which includes accumulating said difference to produce an accumulated number of bits that were available for encoding non-zero AC DCT coefficients in prior 8×

8 blocks of the reduced-quality MPEG coded video but were not used for encoding non-zero AC DCT coefficients in the prior 8×

8 blocks of the reduced-quality MPEG coded video, and making said accumulated number of bits available for encoding non-zero AC DCT coefficients in a certain number of following blocks in the reduced-quality MPEG coded video by dividing said accumulated number of bits by said certain number of following blocks to compute a fraction of the accumulated number of bits that is available for encoding said each of the 8×

8 blocks of said each frame of the reduced-quality MPEG coded video in addition to the number of bits computed by scaling the number of bits encoding non-zero AC DCT coefficients in the corresponding block of the corresponding frame of the original-quality MPEG coded video by the scale factor for said each frame.

4. The method as claimed in claim 3, wherein said certain number of blocks is substantially equal to the number of blocks in said each frame, so that bits that are available but not used for encoding the AC DCT coefficients for the blocks in said each frame are made available for encoding the AC DCT coefficients for the blocks in a following frame.

5. A method of producing reduced-quality MPEG coded video from original-quality MPEG coded video, the original-quality MPEG coded video including a set of non-zero AC discrete cosine transform (DCT) coefficients for 8×

8 blocks in frames of the original-quality MPEG coded video, the reduced-quality MPEG coded video also having frames of 8×

8 blocks, each frame in the reduced-quality MPEG coded video having a corresponding frame in the original-quality MPEG coded video, and each 8×

8 block in each frame of the reduced-quality MPEG coded video having a corresponding block in a corresponding frame in the original-quality MPEG coded video, said method comprising the steps of;

selecting non-zero AC DCT coefficients from corresponding blocks in the original-quality MPEG coded video to be included in the blocks of the reduced-quality MPEG coded video so that each 8×

8 block in each frame of the reduced-quality MPEG coded video has a number of bits encoding non-zero AC DCT coefficients that is generally proportional to the number of bits encoding non-zero AC DCT coefficients for the corresponding 8×

8 block in the corresponding frame of original-quality MPEG coded video;

which includes computing a difference between a number of bits available for encoding the non-zero AC DCT coefficients in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video and the number of bits used for encoding the non-zero AC DCT coefficients retained in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video, and making said difference available for encoding non-zero AC DCT coefficients of following blocks.

6. The method as claimed in claim 5, which includes accumulating said difference to produce an accumulated number of bits that were available for encoding non-zero AC DCT coefficients in prior 8×

8 blocks of the reduced-quality MPEG coded video but were not used for encoding non-zero AC DCT coefficients in the prior 8×

8 blocks of the reduced-quality MPEG coded video, and making said accumulated number of bits available for encoding non-zero AC DCT coefficients in a certain number of following blocks.

7. The method as claimed in claim 6, herein said accumulated number of bits is made available for encoding non-zero AC DCT coefficients in a certain number of following blocks of the reduced-quality MPEG coded video by dividing said accumulated number of bits by said certain number of following blocks of the reduced-quality MPEG coded video to compute a fraction of the accumulated number of bits that is available for encoding non-zero AC DCT coefficients in said each of the 8×

8 blocks of the reduced-quality MPEG coded video in addition to the number of bits computed by scaling the number of bits encoding non-zero AC DCT coefficients in the corresponding block of the corresponding frame of the original-quality MPEG coded video by the scale factor for said each frame.

8. The method as claimed in claim 6, wherein said certain number of following blocks is substantially equal to the number of blocks in a frame, so that bits that are available but not used for encoding the AC DCT coefficients for the blocks in said each frame are made available for encoding the AC DCT coefficients for the blocks in a following frame.

9. A method of producing reduced-quality MPEG coded video from original-quality MPEG coded video, the original-quality MPEG coded video including a set of non-zero AC discrete cosine transform (DCT) coefficients for 8×

8 blocks in frames of the original-quality MPEG coded video, the reduced-quality MPEG coded video also having frames of 8×

8 blocks, each frame in the reduced-quality MPEG coded video having a corresponding frame in the original-quality MPEG coded video, and each 8×

8 block in each frame of the reduced-quality MPEG coded video having a corresponding block in a corresponding frame in the original-quality MPEG coded video, said method comprising, for each block in the reduced-quality MPEG coded video, the steps of;

(a) determining the number of bits used in encoding non-zero AC DCT coefficients in the corresponding block of original-quality MPEG coded video;

(b) computing a number of bits available for encoding AC DCT coefficients in the original-quality MPEG coded video by scaling the number of bits used in encoding non-zero AC DCT coefficients in the corresponding block of original-quality MPEG coded video with a scale factor; and

(c) selecting non-zero AC DCT coefficients in a certain order from the corresponding block in the original-quality MPEG coded video to be included in said each block of the reduced-quality MPEG coded video until the number of bits available for encoding the AC DCT coefficients in the block in the reduced-quality encoded video is not sufficient for encoding, in the block of the reduced-quality MPEG coded video, any more of the AC DCT coefficients in the corresponding block of original-quality MPEG coded video.

10. The method as claimed in claim 9, wherein said order is a parsing order of the non-zero AC DCT coefficients in the corresponding block in the original-quality MPEG coded video.

11. The method as claimed in claim 9, which includes computing the scale factor from a bit rate of the original-quality MPEG coded video and a desired bit rate for the reduced-quality MPEG coded video.

12. The method as claimed in claim 9, which includes computing the scale factor for each frame of the reduced-quality MPEG coded video from a moving average of the size of the corresponding frames in the original-quality MPEG coded video and a desired frame size for the reduced-quality MPEG coded video.

13. The method as claimed in claim 9, which includes computing a difference between the number of bits available for encoding the non-zero AC DCT coefficients in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video and the number of bits used for encoding the non-zero AC DCT coefficients retained in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video, and making available for encoding non-zero AC DCT coefficients of following blocks said difference between the number of bits.

14. The method as claimed in claim 13, which includes accumulating said difference to produce an accumulated number of bits that were available for encoding non-zero AC DCT coefficients in prior 8×

8 blocks of the reduced-quality MPEG coded video but were not used for encoding non-zero AC DCT coefficients in the prior 8×

8 blocks of the reduced-quality MPEG coded video, and making said accumulated number of bits available for encoding non-zero AC DCT coefficients in a certain number of following blocks in the reduced-quality MPEG coded video by dividing said accumulated number of bits by said certain number of following blocks to compute a fraction of the accumulated number of bits that is available for encoding said each of the 8×

8 blocks of said each frame of the reduced-quality MPEG coded video in addition to the number of bits computed by scaling the number of bits encoding non-zero AC DCT coefficients in the corresponding block of the corresponding frame of the original-quality MPEG coded video by the scale factor for said each frame.

15. The method as claimed in claim 14 wherein said certain number of following blocks is substantially equal to the number of blocks in said each frame, so that bits that are available but not used for encoding the AC DCT coefficients for the blocks in said each frame are made available for encoding the AC DCT coefficients for the blocks in a following frame.

16. A method of producing in real-time a stream of reduced-quality MPEG-2 coded video from a source of original-quality MPEG-2 coded video, the original-quality MPEG-2 coded video including a set of non-zero AC discrete cosine transform (DCT) coefficients for 8×

8 blocks in frames of the original-quality MPEG-2 coded video, the reduced-quality MPEG-2 coded video also having frames of 8×

8 blocks, each frame in the reduced-quality MPEG-2 coded video having a corresponding frame in the original-quality MPEG-2 coded video, and each 8×

8 block in each frame of the reduced-quality MPEG-2 coded video having a corresponding block in a corresponding frame in the original-quality MPEG-2 coded video, said method comprising, for each frame in the reduced-quality MPEG-2 coded video, the steps of;

(a) computing a moving average of the size of the corresponding frame in the original-quality MPEG-2 coded video;

(b) computing a scale factor from the moving average of the size of the corresponding frame in the original-quality MPEG-2 coded video and a desired size of said each frame of the reduced-quality MPEG-2 coded video; and

(c) for each 8×

8 block in said each frame;

(i) determining the number of bits used in encoding non-zero AC DCT coefficients in the corresponding block of original-quality MPEG-2 coded video;

(ii) computing a number of bits available for encoding AC DCT coefficients in the original-quality MPEG-2 coded video by scaling the number of bits used in encoding non-zero AC DCT coefficients in the corresponding block of original-quality MPEG-2 coded video with a scale factor, and

(iii) selecting non-zero AC DCT coefficients in a parsing order from the corresponding block in the original-quality MPEG-2 coded video to be included in said each block of the reduced-quality MPEG-2 coded video until the number of bits available for encoding the AC DCT coefficients in the block in the reduced-quality encoded video is not sufficient for encoding, in the block of the reduced-quality MPEG-2 coded video, any more of the AC DCT coefficients in the corresponding block of original-quality MPEG-2 coded video.

17. The method as claimed in claim 16, which includes computing a difference between the number of bits available for encoding the non-zero AC DCT coefficients in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video and the number of bits used for encoding the non-zero AC DCT coefficients retained in said each of the 8×

8 blocks in each frame of the reduced-quality MPEG coded video, and making available for encoding non-zero AC DCT coefficients of following blocks said difference between the number of bits.

18. The method as claimed in claim 17, which includes accumulating said difference to produce an accumulated number of bits that were available for encoding non-zero AC DCT coefficients in prior 8×

8 blocks of the reduced-quality MPEG coded video but were not used for encoding non-zero AC DCT coefficients in the prior 8×

8 blocks of the reduced-quality MPEG coded video, and making said accumulated number of bits available for encoding non-zero AC DCT coefficients in a certain number of following blocks in the reduced-quality MPEG coded video by dividing said accumulated number of bits by said certain number of following blocks to compute a fraction of the accumulated number of bits that is available for encoding said each of the 8×

8 blocks of said each frame of the reduced-quality MPEG coded video in addition to the number of bits computed by scaling the number of bits encoding non-zero AC DCT coefficients in the corresponding block of the corresponding frame of the original-quality MPEG coded video by the scale factor for said each frame.

19. The method as claimed in claim 18, wherein said certain number of blocks is substantially equal to the number of blocks in said each frame, so that bits that are available but not used for encoding the AC DCT coefficients for the blocks in said each frame are made available for encoding the AC DCT coefficients for the blocks in a following frame.