基于图像信息熵的随机数可视化表达

发布时间：2018-04-09 22:23

本文选题：大数据　切入点：信息熵　出处：《湖南大学学报(自然科学版)》2017年04期

【摘要】：为了提高严格受控环境下信息隐写算法的安全性和鲁棒性,基于网络图像大数据提出了一种针对小容量负载的图像零隐写隐秘通信算法.在构建算法所需的完备库时,首先根据网格描述符对图像进行单元熵的提取以形成熵矩阵,然后提取熵矩阵的特征值以降维从而减少计算量,最后根据量化算法将特征值量化为一组随机数以构建完备库.实验结果表明:该算法由于没有改变原始载体图像,所以对于抗击已知原始载体的统计、比对分析和尺度缩放攻击、旋转攻击具有非常好的效果,在尺度缩放和旋转攻击的实验中数据恢复率达到95%以上.在安全级别高、受监控程度严格、容量较小的隐匿通信中,比如对称密码系统密钥的交换,具有重要的应用价值.
[Abstract]:In order to improve the security and robustness of information steganography algorithm in strictly controlled environment, a novel zero-steganography steganography communication algorithm based on network image big data is proposed.When constructing the complete library required by the algorithm, the unit entropy of the image is firstly extracted according to the mesh descriptor to form the entropy matrix, and then the eigenvalue of the entropy matrix is extracted to reduce the dimension and reduce the computation.Finally, according to the quantization algorithm, the eigenvalues are quantized into a set of random numbers to construct a complete library.The experimental results show that the algorithm has a very good effect on resisting the statistics of known original carrier, comparing analysis and scale scaling attack, and rotating attack, because the original carrier image has not been changed.In the experiments of scale scaling and rotation attack, the data recovery rate is over 95%.In the hidden communication with high security level, strict monitoring and small capacity, such as the key exchange of symmetric cryptosystem, it has important application value.
【作者单位】：武汉大学电子信息学院;中国地质大学(武汉)自动化学院;
【基金】：国家自然科学基金资助项目(62161010) National Natural Science Foundation of China(62161010)
【分类号】：TP391.41

【相似文献】