Research On Remote Sensing Image Encryption Algorithm Based On Wavelet Chaotic Neuron With Time Delay

As more and more multimedia information is exchanged on the Internet,the demand for information confidentiality is also expanding.Digital image plays an important role in multimedia interchange.Different from traditional text data,digital image has the characteristics of large amount of information and intuitive.Compared with digital image,remote sensing image has more data and more bands.So it is very important to encrypt digital image and remote sensing image.Based on the wavelet chaotic neuron model,time delay factor and self-feedback are introduced into the wavelet chaotic neuron model,and two new chaotic neuron models are constructed,namely,the wavelet chaotic neuron with time delay and the wavelet self-feedback chaotic neuron with time delay.NIST(National Institute of Standards and Technology)and TestU01 were used to test the randomness of neuron-generated time series and applied to image encryption.The main research work is as follows:(1)The delay in neuron transmission is often ignored,but theory and practice prove that the delay is inevitable.Based on the wavelet chaotic neuron model,the output of the neuron at time(t-1)is delayed to time t,and a new Gauss wavelet chaotic neuron with time delay model is proposed.By means of bifurcation diagram and Lyapunov exponent diagram,five parameters affecting chaos dynamics in Gauss wavelet chaotic neuron with time delay are studied.The sensitivity of the chaotic neuron model to initial value is studied.NIST and TestU01 were used to test the randomness of the chaotic sequence iteratively generated by the neuron model.A color image encryption algorithm based on Gauss wavelet chaotic neuron with time delay is proposed,which combines image scrambling and diffusion.The encryption algorithm is used to simulate the image,and the security analysis of the experimental results is carried out to verify whether the encryption algorithm can resist various attacks.(2)On the basis of Gauss wavelet self-feedback chaotic neuron,and the signal at time t1 is delayed to time t.A chaotic neuron model with time delay self-feedback is proposed.The bifurcation diagram and Lyapunov exponent diagram are used to analyze the chaotic characteristics of Gauss wavelet chaotic neuron with time delay self-feedback and to study the sensitivity of the chaotic neuron model to initial values.NIST and TestU01 were used to analyze the randomness of the chaotic sequence iteratively generated by the neuron model.And it is applied to image encryption algorithm.The encryption algorithm is used to simulate the image,and the security analysis of the experimental results is carried out to verify whether the encryption algorithm can resist various attacks.(3)According to the characteristics of large amount of information in remote sensing image,wavelet packet transform is firstly carried out before remote sensing image is encrypted.The remote sensing image is splitted into small maps with different resolutions to enhance the effect of encryption and reduce the amount of computation.Scramble the processed small graph,restore the scrambled small graph,and carry on diffusion encryption.In the process of scrambled and diffusion encryption,the chaotic sequences generated by wavelet chaotic neuron with time delay and wavelet chaotic neuron with time delay self-feedback are used respectively,and the experimental results of encryption algorithm are analyzed.