This preview shows page 1-2-3-4-5-6-7-8-9-10-11-12-13-14-95-96-97-98-99-100-101-102-103-104-105-106-107-192-193-194-195-196-197-198-199-200-201-202-203-204-205 out of 205 pages.
ForewordIntroduction1 Scope2 References2.1 Identical Recommendations | International Standards2.2 Additional references3 Definitions4 Abbreviations5 Symbols6 General description6.1 Purpose6.2 Codestream6.3 Coding principles7 Encoder requirements8 Decoder requirements8.1 Codestream syntax requirements8.2 Optional file format requirementsAnnex A Codestream syntaxA.1 Headers and marker segmentsA.1.1 Markers and marker segmentsA.1.2 Types of markers and marker segmentsA.1.3 Syntax similarity with ITU-T Rec. T.81 | ISO/IEC 10918-1A.1.4 Marker and marker segment and codestream rulesA.1.5 Key to graphical descriptions (informative)A.2 Information in the marker segmentsA.3 Construction of the codestreamA.4 Delimiting markersA.4.1 Start of codestream (SOC)A.4.2 Start of tile-part (SOT)A.4.3 Start of data (SOD)A.4.4 End of codestream (EOC)A.5 Fixed information marker segmentA.5.1 Image and tile size (SIZ)A.6 Functional marker segmentsA.6.1 Coding style default (COD)A.6.2 Coding style component (COC)A.6.3 Region-of-interest (RGN)A.6.4 Quantization default (QCD)A.6.5 Quantization component (QCC)A.6.6 Progression order change, default (POD)A.7 Pointer marker segmentsA.7.1 Tile-part lengths, main header (TLM)A.7.2 Packet length, main header (PLM)A.7.3 Packets length, tile-part header (PLT)A.7.4 Packed packet headers, main header (PPM)A.7.5 Packed packet headers, tile-part header (PPT)A.8 In bit stream marker segmentsA.8.1 Start of packet (SOP)A.8.2 End of packet header (EPH)A.9 Informational markersA.9.1 Comment and extension (CME)Annex B Data orderingB.1 Image division into componentsB.2 Image division into tiles and tile-componentsB.3 Example of the mapping of components to the reference gridB.4 Tile-component division into resolutions and sub-bandsB.5 Division of resolutions into precinctsB.6 Division of the sub-bands into code-blocksB.7 LayersB.8 PacketsB.9 Packet header information codingB.9.1 Bit stuffing routineB.9.2 Tag treesB.9.3 Zero length packetB.9.4 Code-block inclusionB.9.5 Zero bit-plane informationB.9.6 Number of coding passesB.9.7 Length of the data from a given code-blockB.9.7.1 Single codeword segmentB.9.7.2 Multiple codeword segmentsB.9.8 Order of information within packet headerB.10 Tile Data and Tile-PartsB.11 Progression OrderB.11.1 Progression order determinationB.11.1.1 Layer-resolution-component-position progressiveB.11.1.2 Resolution-layer-component-position progressiveB.11.1.3 Resolution-position-component-layer progressiveB.11.1.4 Position-component-resolution-layer progressiveB.11.1.5 Component-position-resolution-layer progressiveB.11.2 Progression order defaultAnnex C Arithmetic entropy codingC.1 Binary encoding (informative)C.1.1 Recursive interval subdivision (informative)C.1.2 Coding conventions and approximations (informative)C.2 Description of the arithmetic encoder (informative)C.2.1 Encoder code register conventions (informative)C.2.2 Encoding a decision (ENCODE) (informative)C.2.3 Encoding a 1 or a 0 (CODE1 and CODE0) (informative)C.2.4 Encoding an MPS or LPS (CODEMPS and CODELPS)C.2.5 Probability EstimationC.2.6 Renormalization in the encoder (RENORME) (informative)C.2.7 Compressed data output (BYTEOUT) (informative)C.2.8 Initialisation of the encoder (INITENC) (informative)C.2.9 Termination of coding (FLUSH) (informative)C.3 Arithmetic decoding procedureC.3.1 Decoder code register conventionsC.3.2 Decoding a decision (DECODE)C.3.3 Renormalization in the decoder (RENORMD)C.3.4 Compressed data input (BYTEIN)C.3.5 Initialisation of the decoder (INITDEC)C.3.6 Resetting arithmetic coding statisticsC.3.7 Saving arithmetic coding statisticsAnnex D Coefficient bit modelingD.1 Code-block scan pattern within code-blocksD.2 Coefficient bits and significanceD.3 Decoding passes over the bit-planesD.3.1 Significance propagation decoding passD.3.2 Sign bit decodingD.3.3 Magnitude refinement passD.3.4 Cleanup passD.3.5 Example of coding passes and significance propagation (informative)D.4 Initializing and terminatingD.4.1 Decoder terminationD.4.2 Arithmetic encoder terminationD.4.3 Length computation (informative)D.5 Error resilience segmentation symbolD.6 Selective arithmetic decoding bypassD.6.1 Undoing the effects of bit stuffingD.6.2 Predictable terminationD.7 Vertically causal context formationD.8 Flow diagram of the code-block codingAnnex E QuantizationE.1 Scalar coefficient dequantization (normative)E.2 Scalar coefficient quantization (informative)Annex F Discrete wavelet transformation of tile componentsF.1 Introduction and overviewF.1.1 Tile component parametersF.1.2 Discrete Wavelet Transformations (informative)F.1.2.1 Low-pass and high-pass filteringF.1.2.2 Levels of decompositionF.1.2.3 Discrete wavelet filters (informative)F.2 The inverse discrete wavelet transformation (normative)F.2.1 The IDWT procedureF.2.2 The 2D_SR procedureF.2.3 The 2D_INTERLEAVE procedureF.2.4 The HOR_SR procedureF.2.5 The VER_SR procedureF.2.6 The 1D_SR procedureF.2.7 The 1D_EXTR procedureF.2.8 The 1D_IFILTR procedureF.2.8.1 Reversible 1D inverse filteringF.2.8.2 Irreversible 1D inverse filteringF.3 Forward Transformation (informative)F.3.1 The FDWT procedureF.3.2 The 2D_SD procedureF.3.3 The VER_SD procedureF.3.4 The HOR_SD procedureF.3.5 The 2D_DEINTERLEAVE procedureF.3.6 The 1D_SD procedureF.3.7 The 1D_EXTD procedureF.3.8 The 1D_FILTD procedureF.3.8.1 Reversible 1D filteringF.3.8.2 Irreversible 1D filteringAnnex G DC level shifting and component transformationsG.1 DC level shifting of tile componentsG.1.1 DC level shifting of tile components (informative)G.1.2 Inverse DC level shifting of tile components (normative)G.2 Reversible component transformation (RCT)G.2.1 The Forward RCT (informative)G.2.2 The Inverse RCT (normative)G.3 Irreversible component transformation (ICT).G.3.1 The Forward ICT (informative)G.3.2 The Inverse ICT (normative)G.4 Chrominance component sub-sampling and the image reference grid (informative)G.4.1 Interpretation of multiple componentsAnnex H Coding of images with Regions of InterestH.1 Description of the Maxshift methodH.1.1 Encoding (informative)H.1.2 Selection of scaling value, s, at encoder side (normative)H.1.3 Decoding (normative)H.2 Region of interest mask generationH.2.1 Region of Interest mask generation using the W5X3 filter (informative)H.2.2 Region of Interest mask generation using the W9X7 filter (informative)H.3 Remarks on Region of Interest codingH.3.1 Multi-component remarkH.3.2 Disjoint regions remarkH.3.3 Implementation Precision
View Full Document
Unlocking...