摘要
提出了一种基于预测模式相关性的高效视频编码(HEVC)隐写检测方法。首先定义帧内预测模式相关性的概念;然后通过实验,分析隐写前后帧内4×4亮度块中35种帧内预测模式的相关性统计数据的变化情况;最后选择35个隐写前后变化较大的相关性数据构成视频特征,并利用支持向量机(SVM)分类器进行隐写检测。实验表明,本文检测方法的特征维数低,计算复杂度小,对修改帧内预测模式的隐写算法具有良好的检测效果。
At present,information hiding algorithm based on modified intra-prediction mode has been pro- posed in the H. 265/HEVC. The steganographic algorithm embeds secret information by modifying the intra-prediction mode. But there is no steganalysis method for H. 265/HEVC from the available litera- ture. This paper proposes a steganalysis method based on the correlation of intra-predietion mode. The concept of the correlation of intra-prediction mode is defined firstly, then we analyze the correlation of 35 kinds of intra-prediction modes in the 4 X 4 luma block before and after information hiding through abun- dant experimenl data. 35 correlation data which change a lot after information hiding are chosen to con- stitute the video features. Lastly, the video features will be sent to support vector machine (SVM) classi- fier to finish the detection of information hiding. Experimental results demonstrate that the algorithm has good detection effect while the feature dimension is low, and the computational complexity is small.
出处
《光电子.激光》
EI
CAS
CSCD
北大核心
2017年第7期766-772,共7页
Journal of Optoelectronics·Laser
基金
国家自然科学基金(61301247
61300055)
浙江省自然科学基金(Z15F020010)
浙江省重中之重学科开放基金(xkx11405)
宁波大学优秀学位论文培育基金(py2013005)资助项目