高级检索
孙鹏, 郎宇博, 巩家昌, 沈喆. 拼接篡改伪造图像的色彩偏移量不一致取证方法[J]. 计算机辅助设计与图形学学报, 2017, 29(8): 1408-1415.
引用本文: 孙鹏, 郎宇博, 巩家昌, 沈喆. 拼接篡改伪造图像的色彩偏移量不一致取证方法[J]. 计算机辅助设计与图形学学报, 2017, 29(8): 1408-1415.
Sun Peng, Lang Yubo, Gong Jiachang, Shen Zhe. Authentication Method for Splicing Manipulation with Inconsistencies in Color Shift[J]. Journal of Computer-Aided Design & Computer Graphics, 2017, 29(8): 1408-1415.
Citation: Sun Peng, Lang Yubo, Gong Jiachang, Shen Zhe. Authentication Method for Splicing Manipulation with Inconsistencies in Color Shift[J]. Journal of Computer-Aided Design & Computer Graphics, 2017, 29(8): 1408-1415.

拼接篡改伪造图像的色彩偏移量不一致取证方法

Authentication Method for Splicing Manipulation with Inconsistencies in Color Shift

  • 摘要: 拼接篡改是一种常见的伪造图像方法,根据拼接篡改伪造图像中拼接区域与原始区域之间存在的色彩偏移量的差异,提出一种基于偏色估计的拼接篡改伪造图像自动检测方法.首先将图像分为n×n大小的图像子块,利用改进的平均色差计算方法估计每一个子块的色彩偏移量;然后将分块之后的待检验图像的上、左、右3个径向方向上的尺度边缘子块的集合设定为参考区域;计算每一个子块与参考区域之间的偏色距离,最后与设定的偏色距离阈值进行比较后定位图像中的拼接区域,从而揭示拼接篡改图像中存在的色彩偏移量不一致现象.实验结果表明,该方法能够自动检测拼接篡改图像中的色彩偏移量不一致并定位拼接篡改区域,为拼接篡改伪造图像的取证提供了一类科学量化的检验依据.

     

    Abstract: Splicing manipulation is a common way for image forgeries,according to the difference of color shift between splicing areas and original part,we put forward a kind of auto-detection method for splicing images based on estimation of color shift.First of all,the suspicious image is divided into small sub-blocks,size of which are n×n.After that an improved average color deviation calculation method is applied to estimate color shift of each sub-blocks.And then the collection of sub-blocks is set in radial directions of top,left,right on the suspicious image as the reference region,with scale selection of edge sub-blocks,color shift between each sub-block and reference is calculated.At last,the splicing area can be localized automatically by the comparison of color shift distance with empirical threshold.As a result,the inconsistency of color shift in a splicing image can be exposed after the localization of the splicing area.Experiments show that the proposed method can detect splicing images and localize splicing area automatically,and provide a scientific and quantitative evidence for the detection of splicing manipulation.

     

/

返回文章
返回