数字图像拼接篡改盲取证技术研究

Digital image mosaicing tamper with the blind forensic technology research

研究图像的篡改识别问题,由于数字图像能够被轻易的篡改并且很难发现改动痕迹,对篡改像素不融合现象识别不清,导致图像篡改很难被肉眼识别。为解决上述问题,从篡改者的角度对目前流行的篡改手段做了新的分类,详细分析了各种篡改取证技术的优缺点。提出了一种基于统计特征分类的盲检测算法,实验表明,从图像的双谱幅值和相角检测可以准确识别出篡改后的图像,为篡改图像的识别提供了依据。

The image to identify problems tampering, because the digital image can be easily and it is hard to find traces tampering with changes to tamper with the pixels don＇t fusion phenomenon recognition is not clear, cause image is hard to tamper with the naked eye identification. To solve the above problems, from the point of view of the tamper with the popular means tampering with the new classification, and analyzed the advantages and disadvantages of various tamper with the forensic technology. Put forward based on the statistical characteristics of the classification blind detection algorithm, experimental results show that the double spectrum from image amplitude and phase detection can accurately identify tamper with the image, image recognition to tamper with to provide the basis.