拼接篡改伪造图像的色彩偏移量不一致取证方法

发布时间：2018-03-12 17:39

本文选题：拼接篡改　切入点：伪造图像　出处：《计算机辅助设计与图形学学报》2017年08期 　论文类型：期刊论文

【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]:Mosaic tampering is a common method of falsifying images. According to the difference of color offset between the mosaic region and the original region in the mosaic tamper falsification image, In this paper, an automatic detection method of stitching tamper forged image based on color offset estimation is proposed. Firstly, the image is divided into n 脳 n image subblocks, and the color offset of each sub-block is estimated by using the improved average color difference calculation method. Then, the set of scale edge sub-blocks in three radial directions, left and right, is set as the reference area, and the color deviation distance between each sub-block and the reference region is calculated. Finally, after comparing with the set offset distance threshold, the splicing area in the image is located, thus revealing the color offset inconsistency in the mosaic tamper image. The experimental results show that, This method can automatically detect the color offset inconsistency and locate the tampering area in the mosaic tamper image. It provides a scientific and quantitative test basis for the evidence collection of the mosaic tamper falsification image.
【作者单位】：中国刑事警察学院声像资料检验技术系;辽宁网络安全执法协同创新中心;辽宁石油化工大学信息与控制工程学院;中国科学院沈阳自动化研究所机器人学国家重点实验室;
【基金】：国家自然科学基金(61307016) 辽宁省自然科学基金(2014020193) 辽宁省优秀人才支持计划(14021B171210Z)
【分类号】：TP391.41

【参考文献】