基于像素匹配的图像“复制-粘贴”篡改检测算法

“复制-粘贴”是一种简单有效的图像篡改技术，它把图像中某一部分复制后，粘贴到同一图像的另一区域中，以达到伪造图像的目的针对此篡改技术，提出了一种基于像素匹配的篡改检测算法．算法采用离散小波变换（DWT）对待测图像进行降维，对降维后的图像用相位相关求得“复制一粘贴”区域间的位移，根据所求位移采用像素匹配（即将待测图像与移位后的图像求差来定位“复制-粘贴”区域，并利用形态学滤波技术来进一步改善定位效果．实验结果表明，本文算法简便快捷，不仅能够对篡改区域进行准确定位，而且能够有效抵抗多种“复制一粘贴”篡改的后处理操作，包括JPEG、JPEG2000有损压缩及滤波等

An Algorithm for Image Copy-Paste Forgery Detection by Pixel Matching

Copy-paste is an efficient type of image forgery ,in which a part of a digital image is copied and pasted to another part of the same image in order to forge the image. Based on image pixel matching,a new image forensics approach has proposed for detecting copy-paste forgery. First,discrete wavelet transform （DWT）was applied to the input image to yield a reduced dimension representation. Then the phase correlation was computed to estimate the spatial offset between the copied part and the pasted part. The copy-paste regions can be easily located by virtue of pixel matching,which shifted the input image by the spatial offset to calculate the difference between the image and its shifted version. At last,the mathematical morphological operations were used to improve the detection. Experimental results show that the method is easily implemented ;it can achieve an effective and accurate location for tampered areas and is reasonably robust to various types of copypaste post processing such as filtering ,lossy JPEG ,JPEG2000 compression and the combination of these operations.