图像拼接篡改的自动色温距离分类检验方法

拼接篡改是一类常见的图像伪造手段，现有取证方法难以实现图像中拼接篡改区域的自动检测与精确定位，导致拼接篡改伪造图像的取证长期依赖人工经验.基于图像中原始区域与拼接篡改区域所反映的光源色温的差异性，提出一种自动色温距离阈值分类的图像拼接篡改检测与定位方法.首先，变换待检验图像至YCbCr色彩空间，并按照Grid-based方式结构化分解为大小的子图像块；然后，利用自动白平衡（Automatic white balance，AWB）中的白点检测原理对每一个子图像块进行色温估计，计算子图像块与参考区域之间的色温距离；最后，采用最大类间方差法自适应地求取色温距离分类的最佳阈值，对子图像块进行分类标注，实现了图像拼接篡改区域的自动检测与精确定位.实验表明，该方法能够实现图像拼接篡改区域的自动检测与定位，具有较高的量化检测精度.

Detection of Image Splicing Manipulation by Automated Classification of Color Temperature Distance

Splicing is a common types of tampering in image manipulation. As many authentication methods cannot detect and localize the manipulated area automatically in splicing images, authentication of splicing image has depended on human experience for a long time. In this paper, considering the inconsistency of color temperature between original area and splicing area, we propose an automated distance threshold classification method for splicing image detection and manipulation localization by color temperature estimation. At first, we transform suspicious image into YCbCr color space and divide it into blocks with grid-based manner. Then, we estimate color temperature of each block using automatic white balance (AWB) theory, and calculate Euclidean distance between reference area and suspicious area. Finally, we localize the splicing area with an automated estimated optimal threshold of color temperature distance. Experiments indicate that our method can detect splicing images and localize splicing area effectively and automatically with a quantitative result.