一种基于模式噪声熵的图像来源取证算法

针对数字图像的来源取证,提出了一种基于模式噪声熵的检测算法。传感器作为数码相机的重要部件,由于在制造过程中的缺陷,成像时会给图像带来一种模式噪声。该算法利用传感器产生的模式噪声具有唯一性这一特点,对图像进行小波降噪并提取图像的模式噪声,利用模式噪声的熵值对不同来源的图像进行区分。实验结果表明,该方法对原始图像有较高的检测率,对有损压缩图像也有较好的鲁棒性。

New algorithm based on entropy of pattern noise for forensic of image source

Aimed at forensic of digital image source, this paper proposed a detection approach based on entropy of pattern noise. As an important component of digital camera, because its defect in the process of production, there was pattern noise leaved in images. Based on the unique of pattern niose, this arithmetic debased the noise of images using wavelet transform and picked up pattern noise. At last, it distinguished different source of images by the entropy of pattern noise. The experimental results show that the algorithm has a better efficiency with the original images and high robustness with the JPEG compression images.