预览加载中，请您耐心等待几秒...

基于量化的双树复小波变换鲁棒图像水印算法(英文).docx

预览
1/2
2/2

在线预览结束，喜欢就下载吧，查找使用更方便

下载提示 文本预览

如果您无法下载资料，请参考说明：

1、部分资料下载需要金币，请确保您的账户上有足够的金币

2、已购买过的文档，再次下载不重复扣费

3、资料包下载后请先用软件解压，在使用对应软件打开

基于量化的双树复小波变换鲁棒图像水印算法(英文) Introduction: Inrecentyears,theprotectionofdigitalcontenthasbecomeamajorconcernduetotheeaseofillegalreproductionanddistributionofdigitalmedia.Thishasledtothedevelopmentofvariousdigitalwatermarkingtechniquesthatcanbeusedtoprotectdigitalcontentfromunauthorizedaccessorreplication.Thedouble-treecomplexwavelettransform(DT-CWT)hasemergedasapowerfultoolforimageandvideowatermarkingduetoitsabilitytogeneratearobustwatermarkinthetransformdomain.However,theDT-CWTsuffersfromthedisadvantageofhavinglowrobustnesstocommonimageprocessingtechniquessuchascompressionandnoise. Toaddressthischallenge,anovelwatermarkingalgorithmbasedonthequantizationoftheDT-CWTtransformcoefficientsisproposed.Thisalgorithmalsoemploysarobustnesstechniquethroughtheuseofabinarypseudorandomsequencegeneratortogenerateawatermarkthatcanwithstandcommonimageprocessingtechniques. Approach: Theproposedalgorithmisdividedintotwostages;watermarkembeddingandwatermarkextraction.Intheembeddingstage,abinarywatermarkisgeneratedusingapseudorandomsequencegenerator.Thewatermarkisthenembeddedintothelow-frequencycoefficientsoftheDT-CWTtransformbyapplyingquantizationtothesecoefficients.Thequantizationiscarriedoutbycomparingthecoefficientvaluetoathresholdvalueandsettingthecoefficienttoeither0or1dependingonthecomparisonresult. Theextractionstageinvolvestheretrievalofthewatermarkfromthewatermarkedimage.TheextractionprocessutilizesthesameDT-CWTtransformastheembeddingprocesstogeneratetransformcoefficients.Thecoefficientsarethenquantizedusingthesamethresholdvalueusedduringtheembeddingprocess.Thewatermarkisretrievedbycomparingthequantizedcoefficientstothethresholdvalueandgeneratingabinaryoutputaccordingly. Results: Theproposedalgorithmwastestedonasetofgrayscaletestimagesandcomparedwithotherstate-of-the-artDT-CWTbasedwatermarkingalgorithms.Theresultsshowedthattheproposedalgorithmoutperformedtheothersintermsofrobustnessagainstcommonimageprocessingtechniquessuchascompressionandnoise.Thealgorithmwasalsotestedforvariou