一种基于压缩图像的反取证方法
2015-10-21程格平王毅
程格平 王毅

摘 要: 图像处理软件的广泛使用使数字图像的篡改和伪造变得更加容易,这给图像数据的安全性带来严重影响。数字图像取证是解决这个问题的关键技术,逐渐成为研究热点。反取证技术能够有效降低或消除取证技术的检测效果,尚没有得到应有的重视。提出一种针对JEPG压缩的反取证技术,通过在压缩图像的DCT系数中添加适当的噪声移除量化块效应,从而去除图像取证的检测证据。实验结果表明,所提出的方法能够明显降低JEPG图像取证方法的检测性能。
关键词: 压缩图像; 图像取证; 反取证技术; 量化效应; 检测性能
中图分类号:TP391 文献标志码:A 文章编号:1006-8228(2015)09-12-02
Anti-forensic method based on image compression
Cheng Geping, Wang Yi
(School of Mathematical and Computer Sciences, Hubei University of Arts and Science, Xiangyang, Hubei 441053, China)
Abstract: The widespread use of image processing software makes it easy to tamper and counterfeit a digital image, which will lead to serious influence to the security of image data. The digital image forensics is the key technology to solve the problem and is gradually becoming the research focus. Anti-forensics technology can effectively reduce or eliminate the detection effect of the forensics, but it has not been paid due attention. In this paper, an anti-forensics technology is proposed for JEPG compression, which removes the quantization blocking artifacts by adding appropriate noise to the DCT coefficients in a compressed image, so as to eliminate forensic detection evidence of the image. The experimental results show that the proposed method can significantly reduce the detection performance of the JEPG image forensics.
Key words: compressed image; image forensics; anti-forensics; quantization artifacts; detection performance
0 引言
數码相机的迅速普及和多媒体技术的快速发展,使得数字图像的获取、修改和编辑更加简单,同时也对涉及数字图像原始性、真实性和完整性的应用领域带来日趋严重的安全隐患。因此,数字图像的内容保护和真实性鉴定逐渐成为国内外研究的重要课题。图像取证技术通过分析图像数据的统计特性,对伪造或篡改的数字图像进行检测和鉴定,具有广泛的实际应用前景。
近年来,出现了许多有效的图像取证技术和方法。Hany Farid等人[1]利用来源图像在不同方向的高阶统计特性,采用多尺度小波分解和高阶统计建模的方法对照片图像、扫描图像和计算机生成图像进行取证,获得了较好的取证效果,尤其对计算机生成图像具有较高的检测效率。
JEPG是数字图像压缩最为流行的压缩标准,广泛应用于各种图像处理领域。……
