Top Banner
MPEG-1 and MPEG-2 Digital Video Coding Standards Author: Thomas Sikora Presenter: Chaojun Liang
33

MPEG-1 and MPEG-2 Digital Video Coding Standards Author: Thomas Sikora Presenter: Chaojun Liang.

Jan 20, 2016

Download

Documents

Dina Barrett
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
  • MPEG-1 and MPEG-2 Digital Video Coding StandardsAuthor: Thomas SikoraPresenter: Chaojun Liang

  • OverviewIntroductionIntra-picture Coding TechniquesInter-picture Coding Techniques

  • What is MPEG?Stands for Moving Picture Expert GroupNickname to a family of International standards used for coding audio-visual information in a digital compressed format. Includes MPEG-1, MPEG-2, and upcoming MPEG-4.

  • MPEG-1Finalized in 1991Optimized to work at 352*240, 30 f/s or 352*288, 25f/sBit-rate is optimized for application of around 1.5 Mb/secDefined for progressive frames only

  • MPEG-2Finalized in 1994Digital television broadcastingTarget bit-rate raised to 4-9Mb/secMain Profile, Main Level720*480 pixels, 30 frame/sec, 15Mb/secMain Profile, High Level (HDTV)1920*1280 pixels, 30 frame/sec, 80Mb/sec

  • MPEG Video Layers

  • Intra Frame Coding TechniquesVideo FilterDCTQuantizerRun-length EncodingBitstream BufferBit-Rate Control

  • Video FilterSubjective RedundancyEyes more sensitive to changes in Luminance, less to variations in ChrominanceR G B -> Y Cb CrSample Chrominance 2:1 in both directions

  • Video FilterR G BY Cb CrY Cb CrColor Space TransformSample

  • Statistical Redundancy

  • Discrete Cosine TransformClosely related to Discrete Fourier TransformPerformed on 8*8 blockForward DCTInverse DCT

  • DCT Basis Patterns

  • DCT Basis PatternsIncreasing Vertical Frequency (Down)Increase Horizontal Frequency (Right)

  • Variance Distribution of DCT Coefficients

  • DCT Coefficient Quantization

    Higher frequency -> Larger step size -> more coarsely quantizedForce as many of the DCT-Coefficients as possible to zero or near zero

  • Example Coefficients

  • Run-Length EncodingRun-Amplitude PairsBit StreamVariable Length Code Table

  • Inter Frame Coding TechniquesExploit temporal redundancyUse Motion Estimation to eliminate temporal redundancy

  • Picture TypesIntra Pictures (I-Pictures)Predicted Pictures (P-Pictures)Bi-directional Pictures (B-Pictures)

  • Intra PicturesCoded with only info in the picture itselfRandom access points into the compressed dataModerate compress: about 2 bits per coded pixel

  • Predicted PicturesCoded with respect to the nearest previous I-Picture or P-Picture

  • Bi-directional PicturesUse both a past and future picture as a reference.

  • Video Stream CompositionEncoder chose frequency and location of I-PicturesTypical display order

  • Video Stream vs Display order

  • Motion EstimationTree moved down to the rightPeople moved farther to the right than treeFrame 1Frame 2

  • Motion EstimationTo adequately represent the change, or the difference, between 2 video framesFrame-wise direct minus not effectiveMacroblock-wise minus 2-D spatial search for each luminance macroblock

  • Macroblock MatchingFrame 1Frame 2

  • Picture 1Predicted PicturePicture 2Residual Error Picture

  • Coding of Residual ErrorsSimilar to Intra frame codingDifferent quantization matrixMotion vectors are also coded