文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 人文社科 > 教育学/心理学 > Model+classification+rate+control+algorithm+for+video+coding
侵权投诉

Model+classification+rate+control+algorithm+for+video+coding

Journal

of&stems

EngineeringandElectronics,V01.16,No.3,2005,PP.494~497

Model

classificationratecontrol

GuJixing,Zheng

algorithm

forvideo

coding粜

Shibao,WangJia&SunJun

Inst.ofImageCommunication&InformationProcessing,ShanghaiJiaotongUniv.,Shanghai200030,P.R.China

(ReceivedMay25,2004)

Ahstract:Amodelclassification

ing

to

rate

controlmethodforvideocodingis

are

pro恻t.Themacro-blocks

current

are

classifiedaccord—

theirpredictionerrors,anddifferentparametersusedintherate-quantizationanddistortion-quantizationmodel.

Thedifferentmodel

parametersarecalculatedfromtheprevious

frame

ofthesametypeintheprocessofcoding.These

modelsareused

such

to

as

to

estimatetherelations

amongrate,distortionandquantizationofthe

frame.Further

steps,

R—Doptimizationbasedquantizationadjustmentandsmoothingofquantizationofadjacent

can

macroblocks,are

used

improvethequality.Theresultsoftheexperimentsprovethatthetechniqueiseffectiveand

presentedinthepapercanbe

a

berealizedeasily.The

method

goodwayforM匝)EGandH.264

rate

contr01.

Keywords:rate

control,modelclassification,macroblock

ofthesametype,todifferentmacroblockclass.Thenafterfinishingdoingmotioncompensation

to

the

cur—

Ratecontr01is

an

importantissuealthoughitdoes

not

rent

frame(apart

are

fromI

to

frame),quantization

stepsof

belong

to

thenormativepartinvideocoding

rate

stan—

alldassescalculatedmeetthetargetbitsallocated.

dard.Without

tems

can

controlmanyvideocodingsys—

2.NIoDELCLASSIFICATIoNBASED

RATECONTROLTECHNIQUE

2.1

notrun

smoothlybecausetheclientbuffer

a

mayoftenunder-flowandover-flowwhenused

to

channel

deliverthecompressedstreamisofconstant

are

Model

Analysis

bandwidth.Thereseveral

rate

controlwaysfor

L

a

constantbit

rate

video

transmission

system.Onesim—

Itisknown

that.theoreticalrate—distortionanddis—

relationsof

as

a

plesolutionisshowninⅣ皿EG一2TM5lj.anditisused

to

tortion—quantization

Gaussianrandom

variableX

can

be

a

benchmarkofdifferentkindsofnewal—

beexpressed

gorithms.

In~妒EG.2TM5.the

edintothree

R=丢1082面G2

rate

(1)(2)

12for

control

can

bedivid—

parts:frame

leveltargetbitallocation,

D:iQ2

|‘)

where

a2

bufferfullnessbasedquantization

nationandadaptive

parameter

on

determi—

isthevarianceof

X,and母equals

quantization.Some

investigationstheR—D

or

uniform.GaussianandLaplaciandistributionsif仃2ismuch1argerthanQ.

For

focus

on

thebitallocationstep,based

a

R—Qmodelof

whole

frame[2,3].and

to

manyother

quantization

MC-DCT

type

videocodingmethodsuch

error

as

methodspaymoreattention

stepofevery

determineMPEG,theDCreoefficientsofprediction

beconsideredidentical

to

can

macro-block[4,5|.Hsueh—MingHang[6]

a

bezero—meani.i.d.(independent

source.ThustheR—Q

as

fullyexploredtheR—-DandD—.Qrelationofblock.

130

single

distribution)random

a

Tao[7]used

a

similarbut

not

on

thesame

andR—Drelationsofmacroblockarefollows

model,andmadeperceptuallybasedclassification.

Inthispaper,wepropose

a

macroblock

R删=丢薹,啦鑫=

model

classification

to

scheme,whichclassifiesthemacroblocksaccordingtheirpredictionters,which

can

i1哆蕊

N1

o-‘i

(3)

errors.We

assigndifferent

beobtainedfromthe

parame—previousframe

‰=j妻=l警=N磐

DevelopmentProgram0fChina(2002AAl03087)

(4)

采Thisprojectwassuptx_)rtedbytheHi曲TechnologyResearchand

万方数据 

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email