一种图像区域复制篡改的检测算法  

A Robust Detection Algorithm For Image Region-duplication Forgery

作　　者：李峰[1] 刘运松[1] 马婷婷[1] 

机构地区：[1]长沙理工大学计算机与通信工程学院,410114

出　　处：《微计算机信息》2011年第11期9-11,共3页Control & Automation

基　　金：基金申请人:李峰;项目名称:数字图像被动盲取证方法研究;基金颁发部门:国家自然科学基金委(60973113);基金申请人:李峰;项目名称:基于二维EMD的纹理分析方法研究;基金颁发部门:湖南省自然科学基金委(09JJ3120)

摘　　要：本文提出一种基于小波变换和奇异值分解的同幅图"复制-粘贴"篡改检测算法。算法利用离散小波变换提取图像的低频分量,再对低频分量分块提取奇异值特征,用k-d树近邻搜索匹配特征向量,最后以欧式距离作为匹配准则并且配合图像块的坐标偏移信息,进行图像复制粘贴区域的检测和定位。实验结果表明该算法运算量小,检测效率高,不仅能够对篡改区域进行较精确的定位,而且能够有效抵抗压缩、高斯噪声、模糊等多种后处理操作。In this paper, we propose a detection approach for Copy-Move forgery based on wavelet transform and singular value decomposition.This method is implemented by first perform discrete wavelet transform on the image and SVD is used on fixed-size blocks of low-frequency component in wavelet sub-band to yield a reduced dimension representation. Each SV feature vector is then matched to its nearest neighbors using K-D tree in image and copy forgery regions are localized by detecting for all neighborhood vectors.The experiments demonstrate that the proposed approach has low computational complexity and can not only localize the copy forgery regions accurately,but also prove to be robust to post image processing, such as JPEG compression, noise contamination, Gaussian blurring etc.

关 键 词：图像取证 “复制-粘贴”篡改 奇异值分解 块匹配 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]