图像DCT系数的α稳态簇分布模型  

α-Stable Family Models of Image′s DCT Coefficient

作　　者：毛家发[1,2] 林家骏[1] 戴蒙[3] 

机构地区：[1]华东理工大学信息科学与工程学院,上海200237 [2]上饶师范学院数学计算机系,江西上饶334001 [3]上海应用技术学院计算机科学与信息工程系,上海200233

出　　处：《华东理工大学学报（自然科学版）》2009年第3期422-427,共6页Journal of East China University of Science and Technology

基　　金：江西省教育科技基金项目(Gjj08462);上海市教育发展基金会晨光计划(2008CGB21)

摘　　要：提出了一种新颖的统计分布模型———α稳态簇模型。根据图像DCT系数的统计特征,结合理论分析和实验验证,验证了图像DCT系数服从于α稳态簇分布模型。α稳态簇模型具有足够的柔韧性,随着特征指数α改变,其分布形状也会改变。图像纹理细节细密程度与特征指数α呈反比例关系,对于纹理细节细密的图像,其特征指数α偏小;对于具有平滑区域的图像,其特征指数α偏大。A novel statistical distribution model, the α-stable family models of the image＇s DCT coefficients, is proposed. Basing on images DCT coefficients statistical characteristics, together with both theoretical analysis and experimental testifying, it is known that the image DCT coefficients obey the α-stable family distribution model, which has enough flexibility and whose distributing shape changes with the change of characteristic index α. These image textures and details are in inverse to the characteristic index α. The richer the image detail is, the smaller the characteristic index α is.

关 键 词：隐写分析 数字水印 α稳态簇模型 吻合度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]