欢迎来到天天文库
浏览记录
ID:36445212
大小:4.49 MB
页数:122页
时间:2019-05-10
《海量多媒体数字水印技术研究》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、.洳≥:≥-博士学位论文⑧提交日期三墨墨土生二旦浙江大学博士学位论文视觉透明性和稳健性。5.针对目前互联网上广泛流行的MP3音频数据格式,提出了一种鲁棒的音频水印算法,该算法利用音频中节奏和节拍这两个特征,根据音频节拍切分原理对音频进行切分,把水印嵌入在音频高变调区域;利用sNR反向推导理论,根据音频特征动态嵌入水印。具有普遍意义的反向推导理论使得算法具备普遍性,实验结果表明算法具有较高的鲁棒性和抗时间攻击性。6.系统实现。对上述的面向搜索引擎的海量多媒体保护方案,本文给出了详细的系统创建流程,并给出了简
2、单的系统原型实现。关键词数字水印,音频水印,视频水印,多媒体安全,信息安全,海量水印技术,多媒体认证,多媒体检索,网络水印技术。U浙江大学博士学位论文AbstractWithtllefastdevelopmentsofnet、vorkingtec圭lnologiesandmultimediatechnologies,thereisanincreasingdemandofinfonnationsecurityprotectionandmultjmediaautllenticationtechn0109ies
3、.Nowerdays,thecopy—ghtpmtectionisingreatdemaIld,especiallyinthecaseofsearchengine,videoorderingandonlinemusic.Thisthesisnlainlyfocusesonmema四itudewate雠arkinganditsimplementationandthemaincon砸butionsofthisthesisareasfbUows:1.Consid甜ngmei11egaldowmoadingand
4、spreadingofdigitaImedia,afhmeworkwhichworksfinelytoprotectthedi垂talmediacop蜘ghtsundert圭lemassivecircllIllstallce.Thefr锄eworkcombinesmetraditionalcopyri曲tprotectionandsea『chenginetechn0109iesanditconsistsofVideo-audiowateHn酞ingembedding,digitaImediare砸eVal
5、andmediaauthentication.TheVideo—audiowaterIrIa出ngembeddingaIgoritlllniscoⅡ1posedofvideoandaudiowatennarkingembeddingalgorithmswhichcanbewidelyused;boththecontent-basedretrievalandtext-basedretrievalalgorithInsconsititLltemedigitalmediaretrievalalgorithIns
6、;memediaauthenticationmainlyVaIidatesmediawate瑚arkinga11dotherinfomladonrelatedtotheinternet.234Amonitorsystemofmassmultimediaisimplemented.ThegoalofthesystemistomonitorandcorrectthebellaViorofsearchengine.AblindwatenIlarkingtecll士10logyisproposedinmisp印e
7、rtofacilitateVideo—audiowatennarkingembeddingaIldsecurityauthenticationofmeframework.Puttheblemishesof缸aditionalcopyTightprotectiontechn0109iesandt11eprinciplesofzerowatermarkingintoconsideration,anenergyspanfuctionisproposed.Basedonthefunction,awatemlark
8、ingembeddinganddeCectionmodelwhichcandealwhiththemassiVed硒talmediaisproposed.Themodelca工lmakeuseofabundalltinfonnationindigitalmediath锄selvesandexperimentsshowthatthemodelisr。bustandtheperfoHnanceisremarkablewhendea
此文档下载收益归作者所有