基于Krawtchouk不变矩的复制-粘贴篡改盲检测算法  

A Blind Forensic Algorithm for Detecting Copy-Paste Images Based on Krawtchouk Invariant Moments

作　　者：黄子龙[1] 张政保[1] 文家福[1] 刘会英[1] 

机构地区：[1]军械工程学院计算机工程系,河北石家庄050003

出　　处：《计算机技术与发展》2012年第2期168-171,共4页Computer Technology and Development

基　　金：国家自然科学基金项目(60772082);河北省科技基金项目(05213579)

摘　　要：针对数字图像检测中一类常见的复制-粘贴图像篡改,提出一种基于小波变换和径向Krawtchouk不变矩的盲检测算法。算法利用小波变换提取图像的低频分量,对低频分量分块提取径向Krawtchouk不变矩特征,这种特征描述方式对图像旋转后处理具有鲁棒性,然后将特征向量进行按字典排序,并结合数学形态学进行图像复制篡改区域的检测和定位。实验表明该算法能有效地定位出复制和粘贴的图像篡改区域,并对粘贴区域旋转操作具有很强的鲁棒性。For a common class of copy-paste tampering in digital image forensics,a forensics algorithm is proposed based on Discrete Wavelet Transform（DWT） and radial krawtchouk invariance moment.The algorithm uses DWT to extract low frequency components of the image to propose the krawtchouk invariance moment of the blocks of low frequency of components.This characterization method is robust against image rotation.The feature vectors are then sorted according to the dictionary.Combined with mathematical morphology for image temper detection and region localization.Experiments show that the algorithm can locate the tampered region and strong robustness to the rotation of the paste region.

关 键 词：KRAWTCHOUK矩 篡改检测 盲检测 旋转不变性 

分 类 号：TN918[电子电信—通信与信息系统]