基于JPEG图像压缩的痕迹盲取证算法

相机的普及以及图像处理软件的广泛应用,数字图像正面临着被随意篡改和伪造的威胁。针对模糊润饰、拼接篡改操作后的数字图像,提出一种利用块效应和相关性相结合进行定位检测数字图像盲取证方法。数字图像经过模糊润饰、拼接篡改等操作后的区域的块效应和相关性与未篡改区域的块效应和相关性不一致。通过提取块效应特征以及相关性特征来定位篡改区域。实验结果表明,该算法能够有效地对模糊润饰、拼接篡改操作的图像进行检测和定位并且具有很好的鲁棒性。

Blind digital forensics algorithm based on multi-feature

The widely used digital cameras and photo editing software bring us a new problem of tampered image. Focusing on the blurring operation and splicing operation in digital forgeries creation, a new digital image forensic method using blocking artifacts and correlations inconsistency is proposed for localization the tampering region in the digital forgeries. By extracting blocking artifacts and correlations features to locating tampered area. Experimental results show that this method can effectively detect manual blurring and splicing in a tampered image and also has a good robustness.