移动通信
移動通信
이동통신
MOBILE COMMUNICATIONS
2014年
12期
77-82
,共6页
水印检测%Shearlet变换%拉普拉斯分布%柯西分布%广义高斯分布
水印檢測%Shearlet變換%拉普拉斯分佈%柯西分佈%廣義高斯分佈
수인검측%Shearlet변환%랍보랍사분포%가서분포%엄의고사분포
watermark detection%Shearlet transform%Laplace distribution%Cauchy distribution%Generalized gaussian distribution
系统地分析了Shearlet变换系数分布模型，提出了基于Shearlet变换的数字水印算法。将水印检测转化为二元假设检测问题，推导出了水印局域最优非线性检测的通用算法。分别采用拉普拉斯分布、柯西分布和广义高斯分布来对Shearlet变换系数统计分布进行拟合，推导出三种分布下水印检测算法并进行实验。结果显示，与传统的线性相关检测算法相比，这三种检测算法均获得好的检测性能，其中基于广义高斯分布的检测算法性能最佳。
繫統地分析瞭Shearlet變換繫數分佈模型，提齣瞭基于Shearlet變換的數字水印算法。將水印檢測轉化為二元假設檢測問題，推導齣瞭水印跼域最優非線性檢測的通用算法。分彆採用拉普拉斯分佈、柯西分佈和廣義高斯分佈來對Shearlet變換繫數統計分佈進行擬閤，推導齣三種分佈下水印檢測算法併進行實驗。結果顯示，與傳統的線性相關檢測算法相比，這三種檢測算法均穫得好的檢測性能，其中基于廣義高斯分佈的檢測算法性能最佳。
계통지분석료Shearlet변환계수분포모형，제출료기우Shearlet변환적수자수인산법。장수인검측전화위이원가설검측문제，추도출료수인국역최우비선성검측적통용산법。분별채용랍보랍사분포、가서분포화엄의고사분포래대Shearlet변환계수통계분포진행의합，추도출삼충분포하수인검측산법병진행실험。결과현시，여전통적선성상관검측산법상비，저삼충검측산법균획득호적검측성능，기중기우엄의고사분포적검측산법성능최가。
The coefficient distribution model of Shearlet transform is analyzed in detail and a digital watermark algorithm based on Shearlet transform is propose. The watermark detection is converted into binary hypothesis detection, thus a general detection algorithm with local optimal and nonlinear watermark detection is deduced. The statistical distribution of Shearlet coefficients is fitted by using Laplace distribution, Cauchy distribution and generalized Gaussian distribution respectively, thus three watermark detection algorithms are deduced and tested. The results show that the three detection algorithms are better than the traditional linear correlation detection algorithms and the detection algorithm based on generalized Gaussian distribution is the best.
