小波低频子带提取特征的图像复制篡改检测  

Detection of image copy-move forgery based on extracting feature from wavelet sub-band

作　　者：张小祥[1] 周治平[1] 

机构地区：[1]江南大学通信与控制工程学院,江苏无锡214122

出　　处：《计算机工程与应用》2011年第10期161-163,228,共4页Computer Engineering and Applications

摘　　要：针对数字图像取证中一类常见的图像篡改-复制粘贴图像伪造,提出了一种利用小波变换和图像块灰度分布提取特征的检测算法。相对于原图,小波分解的低频子带仍然保持原图像的概貌和空间特性,但在尺寸上减小了很多;对小波低频子带进行重叠分块,再对各重叠块进行灰度分布特征的提取;利用迭代划分法结合相似性匹配搜索相似图像块,进一步减少了检测的计算量;配合图像块的偏移位置信息,进行图像复制伪造区域的检测和定位。实验表明该算法能够较精确地定位出复制和粘贴的图像伪造区域,并有效地减少了运算量,提高了检测效率。Copy-move is a common forgery in the research field of digital image forensics,this paper proposes a detection approach based on wavelet transform and gray distribution feature of image blocks.DWT is applied to the image,and gray distribution feature is extracted on fixed-size blocks of low-frequency component in wavelet sub-band to yield a reduced dimension representation.In order to further reduce the computation,iterative division method combined with the use of similarity are used to search similar image block.Copy forgery regions are localized by detecting for all neighborhood vectors. The experiments demonstrate that the proposed approach can not only localize the copy forgery regions accurately,but also reduce the amount of computation and improve the detection efficiency.

关 键 词：图像取证 小波变换 灰度分布特征 复制伪造区域 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]