基于不变矩特征的图像区域复制粘贴篡改检测  

Detection of image region copy-move forgery based on moment invariant features

作　　者：贺佩[1] 李峰[1] 熊兵[1] 

机构地区：[1]长沙理工大学计算机与通信工程学院,长沙410114

出　　处：《计算机工程与应用》2012年第16期198-201,共4页Computer Engineering and Applications

基　　金：国家自然科学基金(No.60973113);湖南省自然科学基金(No.09JJ3120)

摘　　要：针对一种常见的篡改手段——图像区域复制粘贴,提出了一种基于不变矩特征的检测方法。将图像分成多个重叠块,提取每块的不变矩特征与直方图特征,结合起来得到图像的特征矢量。利用字典排序,依照预定的相似性标准,确定图像中的复制粘贴区域。实验结果表明,该算法在抗旋转操作方面明显优于经典的PCA检测算法,能准确检测出90°和180°的旋转。Targeting a common means of tempering—the image region copy-move forgery,this paper proposes a detection algorithm based on moment invariant features.The detected image is divided into multiple same-sized overlapping blocks,whose moment invariant features and histogram features are extracted to compose a feature vector.The feature vector is sorted by dictionary to locate the copy-move region according to the predetermined similarity criteria.Experimental results show that the proposed algorithm performs better than the classical detection algorithm based on PCA（Principal Component Analysis）in terms of robustness against rotate operation and it can detect the 90° and 180° rotation accurately.

关 键 词：数字图像取证 图像区域复制粘贴篡改 不变矩 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]