摘要
针对遥感技术涉及的数据安全问题,提出了一种小波包变换与混沌神经元结合的遥感图像加密方法。该方法使用SHA-256将初始密钥与GF-2影像相结合,通过小波包变换及自适应分类有效分离GF-2影像的主体信息与不同层次的纹理信息,根据信号分量特征对不同信号进行不同方式的置乱或保留,再进行小波包变换的逆过程,对重构后图像使用改进的混沌神经元动力系统进行像素值扩散。仿真结果表明,该加密方法具有良好的密钥敏感性,充分的密钥空间,经加密后所得各波段密文图像的信息熵、相关系数、直方图等指标均接近理想情况,验证了其能够有效保护遥感图像中的各类信息。
Aiming at the data security of remote sensing technology,a remote sensing image encryption method combining wavelet packet transformation and chaotic neuron is proposed.This method uses SHA-256 to combine the initial key with the GF-2 image,and effectively separates the subject information and the texture information of different levels through wavelet packet transformation and adaptive classification.After the separation,the signals are scrambled or retained in different ways according to the characteristics of signal components.Then,the inverse process of wavelet packet transformation is carried out,and pixel value diffusion is performed on the reconstructed image using the improved chaotic neuron dynamic system.Simulation results show that the encryption method has high key sensitivity and sufficient key space.After encryption,the information entropy,correlation coefficient,histogram and other indicators of the ciphertext images in each band are close to the ideal situation.It is proved that the encryption method can effectively protect all kinds of information in remote sensing images.
作者
徐锡统
陈圣波
于岩
XU Xitong;CHEN Shengbo;YU Yan(College of Geo-exploration Science and Technology,Jilin University,Changchun 130026,China)
出处
《遥感信息》
CSCD
北大核心
2021年第4期76-83,共8页
Remote Sensing Information
基金
吉林省省校共建计划专项(SXGJXX2017-2)
吉林大学高层次科技创新团队建设项目(2017TD-26)。
关键词
GF-2影像
自适应分类
小波包变换
混沌神经元
图像加密
GF-2 image
adaptive classification
wavelet packet transform
chaotic neuron
image encryption
