基于噪声分布规律的伪造图像盲检测算法*

提出了一种基于噪声分布规律的伪造图像盲检测方法。首先利用基于边缘保护的滤波方法检测出图像中的噪声；然后计算图像中同质区域噪声的均值、方差和信噪比等统计量，通过比较图像中同质区域的噪声分布规律的相似性程度实现伪造图像鉴别。实验证明该算法能有效地检测出伪造图像。

Blind detection of digital forgery image based on distribution of noise

This paper proposed a method about blind detection of digital forgery image based on the distribution of noise. Firstly, intended the edge preserving smoothing filtering to get the noise in the image. Then calculated the statistics, such as the gray medium, the variance and the signal noise ratio. Finally, the digital forgery could be detected through comparing the similarity of noise distribution shown by the statistics among the different regions. Experiment results demonstrate the effectiveness of the algorithm in detecting the digital forgery image.