Dirac - A Video Codec

Created by the British Broadcasting Corporation.


Public Member Functions | Protected Member Functions | Protected Attributes | Private Member Functions

dirac::SequenceCompressor Class Reference

Compresses a sequence of frames/fields from a stream. More...

#include <seq_compress.h>

Inheritance diagram for dirac::SequenceCompressor:
Inheritance graph
[legend]
Collaboration diagram for dirac::SequenceCompressor:
Collaboration graph
[legend]

List of all members.

Public Member Functions

Protected Member Functions

Protected Attributes

Private Member Functions


Detailed Description

This class compresses a sequence of frames/fields, frame by frame. or field by field. It currently uses GOP parameters set in the encoder parameters in order to define the temporal prediction structure. A version to incorporate non-GOP structures is TBC.

This is an abstract class.


Constructor & Destructor Documentation

dirac::SequenceCompressor::SequenceCompressor ( StreamPicInput pin,
EncoderParams encp,
DiracByteStream dirac_byte_stream 
)

Creates a sequence compressor, and prepares to begin compressing with the first picture.Sets up picture padding in the picture input if necesary

Parameters:
pinan input stream containing a sequence of frames
encpparameters for the encoding process
dirac_byte_streamOutput destination for compressed data
virtual dirac::SequenceCompressor::~SequenceCompressor ( ) [virtual]

Destructor. Must delete IO objects created by constructor.

dirac::SequenceCompressor::SequenceCompressor ( const SequenceCompressor cpy) [private]

Copy constructor is private and body-less. This class should not be copied.


Member Function Documentation

bool dirac::SequenceCompressor::CanEncode ( ) [protected]
virtual void dirac::SequenceCompressor::CleanBuffers ( ) [protected, virtual]
virtual int dirac::SequenceCompressor::CodedToDisplay ( const int  pnum) [protected, pure virtual]

Uses the GOP parameters to convert picture numbers in coded order to display order. Pure virtual function. The child class will have to define it.

Parameters:
pnumthe picture number in coded order

Implemented in dirac::FrameSequenceCompressor, and dirac::FieldSequenceCompressor.

const EncPicture* dirac::SequenceCompressor::CompressNextPicture ( )

This function codes the next picture in coding order and returns the next picture in display order. In general these will differ, and because of re-ordering there is a delay which needs to be imposed. This creates problems at the start and at the end of the sequence which must be dealt with. At the start we just keep outputting picture 0. At the end you will need to loop for longer to get all the pictures out. It's up to the calling function to do something with the decoded pictures as they come out -- write them to screen or to file, for example. . If coding is fast enough the compressed version could be watched real-time (with suitable buffering in the calling function to account for encode-time variations).

NOTE: LoadNextFrame must be called atleast once before invoking this method.

Returns:
pointer to the next locally decoded picture available for display
DiracByteStats dirac::SequenceCompressor::EndSequence ( )
bool dirac::SequenceCompressor::Finished ( ) [inline]

Indicates whether or not the last picture in the sequence has been compressed.

Returns:
true if last picture has been compressed; false if not

References m_all_done.

const EncPicture* dirac::SequenceCompressor::GetPictureEncoded ( )
virtual bool dirac::SequenceCompressor::LoadNextFrame ( ) [pure virtual]

Load one picture of data into the Sequence Compressor. Sets m_all_done to true if no more data is available to be loaded. Input can be frame or field. So the child class will have to implement this function.

Returns:
true - if frame load succeeded. false - otherwise

Implemented in dirac::FrameSequenceCompressor, and dirac::FieldSequenceCompressor.

void dirac::SequenceCompressor::MakeSequenceReport ( ) [protected]
SequenceCompressor& dirac::SequenceCompressor::operator= ( const SequenceCompressor rhs) [private]

Assignment = is private and body-less. This class should not be assigned..

int dirac::SequenceCompressor::PTSOffset ( ) [inline]

References m_delay.

void dirac::SequenceCompressor::SetMotionParameters ( ) [protected]
virtual void dirac::SequenceCompressor::SetPicTypeAndRefs ( PictureParams pparams) [pure virtual]
void dirac::SequenceCompressor::SignalEOS ( ) [inline]

References m_eos_signalled.

virtual void dirac::SequenceCompressor::UpdateCBRModel ( EncPicture my_picture,
const PictureByteIO picture_byteio 
) [protected, pure virtual]
void dirac::SequenceCompressor::UpdateIntraPicCBRModel ( const PictureParams ,
const bool  is_a_cut 
) [protected]

Member Data Documentation

Referenced by Finished().

Referenced by PTSOffset().

Referenced by SignalEOS().

Flag which is false if we've been all-done for more than one picture, true otherwise (so that we can take actions on finishing once only).


The documentation for this class was generated from the following file:

© 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.