摘要
在数字图像取证领域,如何高效甄别图像是否被篡改已经成为相关科研工作者的研究重点。笔者介绍了一直双压缩图像的检测方法,从特征融合的角度出发,在已有的篡改检测方法中加入特征融合和特征降维的步骤,并引入机器学习方法,依据统计模型来得到最优的检测率。实验表明,基于双压缩的数字图像取证准确率更高,具有更强的鲁棒性。
In the field of digital image forensics,how to effectively identify whether the image has been tampered has become the research focus of relevant researchers.From the point of view of feature fusion,the steps of feature fusion and feature dimensionality reduction are added to the existing tamper detection methods,and machine learning method is introduced to get the optimal detection rate according to the statistical model.Experiments show that the digital image forensics based on double compression has higher accuracy and stronger robustness.
作者
朱琳
Zhu Lin(Wuhan Qingchuan University,Wuhan Hubei 430204,China)
出处
《信息与电脑》
2020年第5期55-57,共3页
Information & Computer
关键词
数字图像取证
双压缩取证
特征融合
PCA
digital image forensics
dual compression forensics
feature fusion
PCA