小波变换与CNN相结合的视频加密算法  

Video EncryptionAlgorithm Combining Wavelet Transform and CNN

作　　者：朱艳平 ZHU Yan-ping(Information Engineering College/Xinyang College of Agriculture and Forestry,Xinyang 464000,China)

机构地区：[1]信阳农林学院信息工程学院,河南信阳464000

出　　处：《山东农业大学学报（自然科学版）》2020年第6期1150-1156,共7页Journal of Shandong Agricultural University：Natural Science Edition

基　　金：河南省科技厅科技攻关计划项目(172102210451);信阳农林学院2018年度优秀基层教学组织建设项目(10)。

摘　　要：针对加密后视频数据量巨大,安全性和实时性难以平衡等问题,本文提出小波变换与细胞神经网络相结合的视频加密算法。该算法采用五维细胞神经网络超混沌系统作为密钥源,初始值和加密密钥的选取,均与明文视频有关;首先对明文进行一级离散小波变换,再对变换后的低频部分进行置乱加密和替代加密,使需要处理的数据量和所需产生的密钥序列呈数量级地减少;对明文进行一阶离散小波变换分解,然后对变换的低频部分进行置乱加密和替换加密,使待处理的数据量和要生成的密钥序列减少一个数量级。置乱加密又分为行置乱、列置乱和像素置乱,替代加密与明文以及前一个像素的加密值相关,提高了安全性。该算法的密钥敏感性和明文敏感性较强,密钥空间大,统计特性完全被打破,具有一定的抗噪性和抗剪裁性,加密后的视频数据量小,加解密效率高,实用性强。In view of the huge amount of encrypted video data,the difficulty in balancing security and real-time,a video encryption algorithm based on wavelet transform and cellular neural network is proposed.Five-dimensional cellular neural network hyper chaotic system is used as the key source,the selection of initial value and encryption key is related to plaintext video.Firstly,the plaintext is decomposed by first-order discrete wavelet transform,then the low-frequency part of the transform is scrambled encryption and replaced encryption,so that the amount of data to be processed and the key sequence to be generated are reduced by an order of magnitude.Scrambling encryption is divided into row scrambling,column scrambling and pixel scrambling,alternative encryption is related to plaintext and the encryption value of the previous pixel,the security of the algorithm is improved.The algorithm has strong key sensitivity and plaintext sensitivity,large key space,completely broken statistical characteristics,and the certain anti-noise and anti-clipping properties.Encrypted video data is small,encryption and decryption efficiency is high,and the practicability is strong.

关 键 词：小波变换 低频 细胞神经网络 视频加密 视频大小 

分 类 号：TP309.7[自动化与计算机技术—计算机系统结构]