随着图像处理技术的快速地发展和广泛地应用, 图像篡改变得越来越容易和普及。恶意篡改的图像会对新闻、司法等领域造成极大的负面影响。对基于高阶统计量双相干理论的图像篡改检测方法进行了深入研究, 论述了该方法的原理, 并应用该方法进行了大量的分类真假图像对照实验, 分析了该方法的适用条件和局限性, 实验结果表明, 该方法的识别准确率与拼接边缘灰度的差异有关。

With the rapid development and wide application of digital image processing, images become more and more easy and wide to be tampered. The abuse of tampered images often appear in domains of news and judicial and did great harm. The higher order statistics biocoherent detecting method is studied thoroughly. A large number of experiments are done. The condition and limitations of the using of this method is analyzed. The experimental results show that the accuracy of the identification has a direct bearing of the differences of the gray value on the edge.