Dirac - A Video Codec

Created by the British Broadcasting Corporation.

dirac Namespace Reference

Classes

class dirac::ArithCodec< T >
Abstract binary arithmetic coding class. More...
class dirac::ArithCodec< T >::Triple
A class for encapsulating interval fractions for use in arithmetic coding. More...
class dirac::ArithCodec< T >::Context
A class for binary contexts. More...
class dirac::Range
Range type. More...
class dirac::OneDArray< T >
A template class for one-dimensional arrays. More...
class dirac::TwoDArray< T >
A template class for two-dimensional arrays. More...
class dirac::BandCodec
A general class for coding and decoding wavelet subband data. More...
class dirac::LFBandCodec
A class specially for coding the LF subbands. More...
class dirac::IntraDCBandCodec
A class specially for coding the DC subband of Intra frames. More...
class dirac::BasicOutputManager
Class for managing bit- and byte-oriented output. More...
class dirac::UnitOutputManager
A class for handling data output, including headers. More...
class dirac::FrameOutputManager
class dirac::SequenceOutputManager
class dirac::BitInputManager
A class for managing bit-wise and byte-wise input. More...
class dirac::CommandLine
struct dirac::CommandLine::option
class dirac::PicArray
A class for picture component data. More...
class dirac::CostType
A structure for recording costs, particularly in quantisation. More...
class dirac::EntropyCorrector
A class used for correcting estimates of entropy. More...
class dirac::OLBParams
Parameters for overlapped block motion compensation. More...
class dirac::SeqParams
Parameters relating to the video sequence being encoded/decoded. More...
class dirac::FrameParams
Parameters for initialising frame class objects. More...
class dirac::CodecParams
Parameters common to coder and decoder operation. More...
class dirac::EncoderParams
Parameters for the encoding process. More...
class dirac::DecoderParams
Parameters for the decoding process. More...
class dirac::QuantiserLists
Class for encapsulating quantiser data. More...
class dirac::Frame
A class for encapsulating all the data relating to a frame. More...
class dirac::FrameBuffer
Holds frames both for reference and to overcome reordering delay. More...
class dirac::MotionCompensator
Motion compensator class. More...
class dirac::MotionVector< T >
Motion vector class - just a pair. More...
class dirac::MvCostData
Class for recording costs derived in motion estimation. More...
class dirac::MvData
Class for all the motion vector data. More...
class dirac::MEData
Class for all the motion estimation data. More...
class dirac::MvDataCodec
Codes and decodes all the Motion Vector data. More...
class dirac::StreamPicOutput
Class for outputting pictures. More...
class dirac::MemoryStreamOutput
class dirac::MemoryStreamOutput::OutputMemoryBuffer
local memory buffer More...
class dirac::FileStreamOutput
class dirac::StreamPicInput
Picture input class. More...
class dirac::MemoryStreamInput
class dirac::MemoryStreamInput::InputMemoryBuffer
Class that defines the Input Stream Memory Buffer. More...
class dirac::FileStreamInput
Picture input class. More...
class dirac::UpConverter
Upconversion class. More...
class dirac::CodeBlock
Class for encapsulating metadata concerning a block of coefficients in a subband. More...
class dirac::Subband
Class encapsulating all the metadata relating to a wavelet subband. More...
class dirac::SubbandList
A class encapulating all the subbands produced by a transform. More...
class dirac::PredictStep< gain >
Class to do two-tap prediction lifting step. More...
class dirac::UpdateStep< gain >
Class to do two-tap updating lifting step. More...
class dirac::WaveletTransform
A class to do wavelet transforms. More...
class dirac::CompCompressor
Compress a frame component. More...
class dirac::FrameCompressor
Compress a single image frame. More...
class dirac::QualityMonitor
Class to monitor the quality of pictures and adjust coding parameters appropriately. More...
class dirac::QuantChooser
Choose a quantiser. More...
class dirac::SequenceCompressor
Compresses a sequence of frames from a stream. More...
class dirac::CompDecompressor
Compress a frame component. More...
class dirac::InputStreamBuffer
Input Stream Buffer Class. More...
class dirac::DiracParser
Dirac Stream Parser Class. More...
class dirac::FrameDecompressor
Compress a single image frame. More...
class dirac::SequenceDecompressor
Decompresses a sequence of frames from a stream. More...
class dirac::BlockMatcher
Class to do block matching. More...
class dirac::DownConverter
A class for fast downconversion of picture data. More...
class dirac::ModeDecider
Decides between macroblock and block prediction modes. More...
class dirac::SubpelRefine
The SubpelRefine class takes pixel-accurate motion vectors and refines them to 1/8-pixel accuracy. More...
class dirac::BlockDiffParams
A class encapsulating parameters for calculating a block difference value (a single instance of matching). More...
class dirac::BlockDiff
An abstract class for doing block difference calculations. More...
class dirac::SimpleBlockDiff
A class for doing block differences without bounds-checking, inherited from BlockDiff. More...
class dirac::BChkBlockDiff
A class for doing block differences with bounds-checking, inherited from BlockDiff. More...
class dirac::IntraBlockDiff
A class for calculating the difference between a block and its DC value (average). More...
class dirac::BiBlockDiff
A virtual class for bi-directional differences. More...
class dirac::BiSimpleBlockDiff
A class for bi-directional differences with two references, and no bounds checking. More...
class dirac::BiBChkBlockDiff
A class for bi-directional differences with two references, with bounds checking. More...
class dirac::BlockDiffUp
An abstract class for doing block differences with an upconverted reference. More...
class dirac::SimpleBlockDiffUp
A class for doing block differences without bounds-checking with upconverted references, inherited from BlockDiffUp. More...
class dirac::BChkBlockDiffUp
A class for doing block differences with bounds-checking with upconverted references, inherited from BlockDiffUp. More...
class dirac::BiBlockDiffUp
An abstract class for doing block differences with two upconverted references. More...
class dirac::BiSimpleBlockDiffUp
A class for doing bi-directional block differences without bounds checking. More...
class dirac::BiBChkBlockDiffUp
A class for doing bi-directional block differences with bounds checking. More...
class dirac::MotionEstimator
Class to handle the whole motion estimation process. More...
class dirac::PixelMatcher

Typedefs

typedef short ValueType
typedef int CalcValueType
typedef MotionVector< int > MVector
MVector class is a vector of ints.
typedef MotionVector< int > ImageCoords
ImageCoords class is a vector of ints.
typedef TwoDArray< MVector > MvArray
MvArray is a two-D array of MVectors.
typedef TwoDArray< MotionVector<
float > > MvFloatArray
An array of float-based motion vectors for doing global motion calcs.
typedef std::vector< std::vector<
MVector > > CandidateList

Enumerations

enum PredMode { INTRA, REF1_ONLY, REF2_ONLY, REF1AND2 }
Prediction modes for blocks. More...
enum CompSort {
Y_COMP, U_COMP, V_COMP, R_COMP,
G_COMP, B_COMP
}
Types of picture component. More...
enum AddOrSub { ADD, SUBTRACT }
Addition or subtraction. More...
enum Direction { FORWARD, BACKWARD }
Forward or backward. More...
enum WltFilter { DAUB, HAAR }
Currently only Daubechies (DAUB) implemented. More...
enum CtxAliases {
  SIGN0_CTX, SIGN_POS_CTX, SIGN_NEG_CTX, Z_BIN1z_CTX,
  Z_BIN1nz_CTX, Z_BIN2_CTX, Z_BIN3_CTX, Z_BIN4_CTX,
  Z_BIN5plus_CTX, NZ_BIN1z_CTX, NZ_BIN1a_CTX, NZ_BIN1b_CTX,
  NZ_BIN2_CTX, NZ_BIN3_CTX, NZ_BIN4_CTX, NZ_BIN5plus_CTX,
  BLOCK_SKIP_CTX
}
Contexts used for coefficient coding. More...
enum MvCtxAliases {
  YDC_BIN1_CTX, YDC_BIN2plus_CTX, YDC_SIGN0_CTX, UDC_BIN1_CTX,
  UDC_BIN2plus_CTX, UDC_SIGN0_CTX, VDC_BIN1_CTX, VDC_BIN2plus_CTX,
  VDC_SIGN0_CTX, REF1x_BIN1_CTX, REF1x_BIN2_CTX, REF1x_BIN3_CTX,
  REF1x_BIN4_CTX, REF1x_BIN5plus_CTX, REF1x_SIGN0_CTX, REF1x_SIGNP_CTX,
  REF1x_SIGNN_CTX, REF1y_BIN1_CTX, REF1y_BIN2_CTX, REF1y_BIN3_CTX,
  REF1y_BIN4_CTX, REF1y_BIN5plus_CTX, REF1y_SIGN0_CTX, REF1y_SIGNP_CTX,
  REF1y_SIGNN_CTX, REF2x_BIN1_CTX, REF2x_BIN2_CTX, REF2x_BIN3_CTX,
  REF2x_BIN4_CTX, REF2x_BIN5plus_CTX, REF2x_SIGN0_CTX, REF2x_SIGNP_CTX,
  REF2x_SIGNN_CTX, REF2y_BIN1_CTX, REF2y_BIN2_CTX, REF2y_BIN3_CTX,
  REF2y_BIN4_CTX, REF2y_BIN5plus_CTX, REF2y_SIGN0_CTX, REF2y_SIGNP_CTX,
  REF2y_SIGNN_CTX, PMODE_BIT0_CTX, PMODE_BIT1_CTX, MB_CMODE_CTX,
  MB_SPLIT_BIN1_CTX, MB_SPLIT_BIN2_CTX
}
Contexts used for MV data coding. More...

Functions

template<class T> std::ostream & operator<< (std::ostream &stream, TwoDArray< T > &array)
A function for extracting array data.
template<class T> std::istream & operator>> (std::istream &stream, TwoDArray< T > &array)
A function for inserting array data.
ValueType BChk (const ValueType &num, const ValueType &max)
A simple bounds checking function, very useful in a number of places.
void dirac_assert (const char *p_fname, int line_number, const char *p_mess)
void dirac_report (const char *p_fname, int line_number, const char *p_mess)
void UnsignedGolombCode (BasicOutputManager &bitman, const unsigned int val)
Code a value using unsigned exp-Golomb coding and output it.
void UnsignedGolombCode (std::vector< bool > &bitvec, const unsigned int val)
Code a value using unsigned exp-Golomb coding and output it to a vector.
void GolombCode (BasicOutputManager &bitman, const int val)
Code a value using signed exp-Golomb coding and output it.
void GolombCode (std::vector< bool > &bitvec, const int val)
Code a value using signed exp-Golomb coding and output it to a vector.
unsigned int UnsignedGolombDecode (BitInputManager &bitman)
Decode a value using unsigned exp-Golomb decoding and output it.
unsigned int UnsignedGolombDecode (const std::vector< bool > &bitvec)
Decode a value using unsigned exp-Golomb decoding and output it.
int GolombDecode (BitInputManager &bitman)
Decode a value using signed exp-Golomb decoding and output it.
int GolombDecode (const std::vector< bool > &bitvec)
Decode a value using signed exp-Golomb decoding and output it.
template<class T> std::ostream & operator<< (std::ostream &stream, MotionVector< T > &mv)
Overloaded operator<< for MotionVector class for output to stream.
template<class T> std::istream & operator>> (std::istream &stream, MotionVector< T > &mv)
Overloaded operator>> for MotionVector class for input from stream.
MVector MvMedian (const MVector &mv1, const MVector &mv2, const MVector &mv3)
Return the median of three motion vectors.
MVector MvMedian (const std::vector< MVector > &vect_list)
Return the median of a set of motion vectors.
MVector MvMean (MVector &mv1, MVector &mv2)
Return the mean of two motion vectors.
int Norm2 (const MVector &mv)
Return the squared length of a motion vector.
int Norm1 (const MVector &mv)
Return the sum of the lengths of a motion vector's componets.
int GetMean (std::vector< int > &values)
Return the mean of a set of integer values.
unsigned int GetMean (std::vector< unsigned int > &values)
Return the mean of a set of unsigned integer values.
void AddNewVlist (CandidateList &vect_list, const MVector &mv, const int xr, const int yr)
Add a new motion vector list of neighbours of a vector to the set of lists.
void AddNewVlist (CandidateList &vect_list, const MVector &mv, const int xr, const int yr, const int step)
Add a new motion vector list to the set of lists for sub-pixel matching.
void AddNewVlistD (CandidateList &vect_list, const MVector &mv, const int xr, const int yr)
Add a new motion vector list of diagnonal neighbours of a vector to the set of lists.
void AddVect (CandidateList &vect_list, const MVector &mv, const int list_num)
Add a motion vector to the set of motion vector lists.
ValueType GetVar (const MVector &mv1, const MVector &mv2)
Get the (absolute) variation between two motion vectors.
ValueType GetVar (const std::vector< MVector > &pred_list, const MVector &mv)
Get the (absolute) variation between a motion vector and a list of motion vectors.

Variables

const unsigned int START_CODE_PREFIX = 0x42424344
Prefix for all start codes.
const unsigned int START_CODE_PREFIX_BYTE0
const unsigned int START_CODE_PREFIX_BYTE1
const unsigned int START_CODE_PREFIX_BYTE2
const unsigned int START_CODE_PREFIX_BYTE3
const unsigned char RAP_START_CODE = 0xD7
Random Access Point (RAP) Intra Frame start Code.
const unsigned char IFRAME_START_CODE = 0xD6
Non-RAP Intra Frame start code.
const unsigned char L1FRAME_START_CODE = 0xD4
L1 Frame start code.
const unsigned char L2FRAME_START_CODE = 0xD5
L2 Frame start code.
const unsigned char SEQ_END_CODE = 0xD0
Sequence end code.
const unsigned char NOT_START_CODE = 0xFF
Not a start code but part of data.
const unsigned char BITSTREAM_VERSION = 0x03
Bitstream version.
const QuantiserLists dirac_quantiser_lists
A constant list of the quantisers being used in Dirac.

Typedef Documentation

typedef int dirac::CalcValueType

typedef std::vector< std::vector< MVector > > dirac::CandidateList

typedef MotionVector<int> dirac::ImageCoords

typedef TwoDArray<MVector> dirac::MvArray

typedef MotionVector<int> dirac::MVector

typedef TwoDArray< MotionVector<float> > dirac::MvFloatArray

typedef short dirac::ValueType

Enumeration Type Documentation

enum AddOrSub

Enumeration values:

ADD

SUBTRACT

enum CompSort

Enumeration values:

Y_COMP

U_COMP

V_COMP

R_COMP

G_COMP

B_COMP

enum CtxAliases

Enumeration values:

SIGN0_CTX

SIGN_POS_CTX

SIGN_NEG_CTX

Z_BIN1z_CTX

Z_BIN1nz_CTX

Z_BIN2_CTX

Z_BIN3_CTX

Z_BIN4_CTX

Z_BIN5plus_CTX

NZ_BIN1z_CTX

NZ_BIN1a_CTX

NZ_BIN1b_CTX

NZ_BIN2_CTX

NZ_BIN3_CTX

NZ_BIN4_CTX

NZ_BIN5plus_CTX

BLOCK_SKIP_CTX

enum Direction

Enumeration values:

FORWARD

BACKWARD

enum MvCtxAliases

Enumeration values:

YDC_BIN1_CTX

YDC_BIN2plus_CTX

YDC_SIGN0_CTX

UDC_BIN1_CTX

UDC_BIN2plus_CTX

UDC_SIGN0_CTX

VDC_BIN1_CTX

VDC_BIN2plus_CTX

VDC_SIGN0_CTX

REF1x_BIN1_CTX

REF1x_BIN2_CTX

REF1x_BIN3_CTX

REF1x_BIN4_CTX

REF1x_BIN5plus_CTX

REF1x_SIGN0_CTX

REF1x_SIGNP_CTX

REF1x_SIGNN_CTX

REF1y_BIN1_CTX

REF1y_BIN2_CTX

REF1y_BIN3_CTX

REF1y_BIN4_CTX

REF1y_BIN5plus_CTX

REF1y_SIGN0_CTX

REF1y_SIGNP_CTX

REF1y_SIGNN_CTX

REF2x_BIN1_CTX

REF2x_BIN2_CTX

REF2x_BIN3_CTX

REF2x_BIN4_CTX

REF2x_BIN5plus_CTX

REF2x_SIGN0_CTX

REF2x_SIGNP_CTX

REF2x_SIGNN_CTX

REF2y_BIN1_CTX

REF2y_BIN2_CTX

REF2y_BIN3_CTX

REF2y_BIN4_CTX

REF2y_BIN5plus_CTX

REF2y_SIGN0_CTX

REF2y_SIGNP_CTX

REF2y_SIGNN_CTX

PMODE_BIT0_CTX

PMODE_BIT1_CTX

MB_CMODE_CTX

MB_SPLIT_BIN1_CTX

MB_SPLIT_BIN2_CTX

enum PredMode

Enumeration values:

INTRA

REF1_ONLY

REF2_ONLY

REF1AND2

enum WltFilter

Enumeration values:

DAUB

HAAR

Function Documentation

void AddNewVlist ( CandidateList & vect_list,

const MVector & mv,

const int xr,

const int yr,

const int step

)

void AddNewVlist ( CandidateList & vect_list,

const MVector & mv,

const int xr,

const int yr

)

void AddNewVlistD ( CandidateList & vect_list,

const MVector & mv,

const int xr,

const int yr

)

void AddVect ( CandidateList & vect_list,

const MVector & mv,

const int list_num

)

Add a motion vector to the set of motion vector lists, making sure it's not a duplicate.

ValueType BChk ( const ValueType & num,

const ValueType & max

) [inline]

void dirac_assert ( const char * p_fname,

int line_number,

const char * p_mess

)

Print a message to standard error and abort if in debug mode

void dirac_report ( const char * p_fname,

int line_number,

const char * p_mess

)

Print a message to standard error

unsigned int GetMean ( std::vector< unsigned int > & values ) [inline]

int GetMean ( std::vector< int > & values ) [inline]

ValueType GetVar ( const std::vector< MVector > & pred_list,

const MVector & mv

)

Return the variation between a motion vector and a list of motion vectos, computed as the sum of absolute differences between the components of the vector and the median vector produced by the list of vectors

ValueType GetVar ( const MVector & mv1,

const MVector & mv2

)

Return the variation between two motion vectors, computed as the sum of absolute differences of their components.

void GolombCode ( std::vector< bool > & bitvec,

const int val

)

void GolombCode ( BasicOutputManager & bitman,

const int val

)

int GolombDecode ( const std::vector< bool > & bitvec )

int GolombDecode ( BitInputManager & bitman )

MVector MvMean ( MVector & mv1,

MVector & mv2

) [inline]

MVector MvMedian ( const std::vector< MVector > & vect_list ) [inline]

MVector MvMedian ( const MVector & mv1,

const MVector & mv2,

const MVector & mv3

) [inline]

int Norm1 ( const MVector & mv ) [inline]

int Norm2 ( const MVector & mv ) [inline]

template<class T>

std::ostream& operator<< ( std::ostream & stream,

MotionVector< T > & mv

)

template<class T>

std::ostream& operator<< ( std::ostream & stream,

TwoDArray< T > & array

)

template<class T>

std::istream& operator>> ( std::istream & stream,

MotionVector< T > & mv

)

template<class T>

std::istream& operator>> ( std::istream & stream,

TwoDArray< T > & array

)

void UnsignedGolombCode ( std::vector< bool > & bitvec,

const unsigned int val

)

void UnsignedGolombCode ( BasicOutputManager & bitman,

const unsigned int val

)

unsigned int UnsignedGolombDecode ( const std::vector< bool > & bitvec )

unsigned int UnsignedGolombDecode ( BitInputManager & bitman )

Variable Documentation

const unsigned char dirac::BITSTREAM_VERSION = 0x03

const QuantiserLists dirac::dirac_quantiser_lists

const unsigned char dirac::IFRAME_START_CODE = 0xD6

const unsigned char dirac::L1FRAME_START_CODE = 0xD4

const unsigned char dirac::L2FRAME_START_CODE = 0xD5

const unsigned char dirac::NOT_START_CODE = 0xFF

const unsigned char dirac::RAP_START_CODE = 0xD7

const unsigned char dirac::SEQ_END_CODE = 0xD0

const unsigned int dirac::START_CODE_PREFIX = 0x42424344

const unsigned int dirac::START_CODE_PREFIX_BYTE0

Initial value:
(START_CODE_PREFIX >> 24) & 0xFF

const unsigned int dirac::START_CODE_PREFIX_BYTE1

Initial value:
(START_CODE_PREFIX >> 16) & 0xFF

const unsigned int dirac::START_CODE_PREFIX_BYTE2

Initial value:
(START_CODE_PREFIX >> 8) & 0xFF

const unsigned int dirac::START_CODE_PREFIX_BYTE3

Initial value:
START_CODE_PREFIX & 0xFF

© 2004 British Broadcasting Corporation. Dirac code licensed under the Mozilla Public License (MPL) Version 1.1.
HTML documentation generated by Dimitri van Heesch's excellent Doxygen tool.