With the rapid development of Internet technology and multimedia communication technology,image and video processing technologies has been widely used in various fields.The characteristics of the Internet itself,such as openness and sharing,make the image and video information be vulnerable to stealing and tampering by criminals in the process of storage and transmission.Therefore,the security of image and video information has received widespread attention and concern.Currently,video encryption technology has become an effective means of information security protection.In practical applications,the complexity and real-time characteristics of video information make the encryption effect of some commonly used algorithms have some shortcomings.Therefore,it is necessary to design practical encryption algorithm based on the specific characteristics of the video.In order to balance high security and low computational complexity,and to ensure the compatibility of the encrypted video stream and the invariance of the compression ratio,this paper studies video encryption technology with good codestream compatibility for the entropy coding stage in the latest video coding standard HEVC.Considering the close relationship between video encryption and image encryption,based on the studying of the chaotic randomness of a class of time-varying bilateral symbol systems,an image chaotic encryption scheme and a video chaotic encryption scheme are proposed respectively.The effect of the new scheme on image and video encryption is verified by computer simulation.The innovative research content of this paper includes the following three parts:Firstly,the Devaney chaos of a class of time-varying bilateral chaotic symbol systems is theoretically proved,and some stochastic performance detections are carried out for the chaotic sequences generated by that time-varying bilateral chaotic symbol systems.The simulation results show that the chaotic sequence has good randomness.Secondly,a chaotic image encryption scheme based on scrambling and diffusion structure is proposed.The scheme firstly uses the breadth-first search method to scramble the pixel position of the original image,and then uses the pseudo-random sequence generated by the time-varying bilateral chaotic symbol system to perform bit-by-bit XOR encryption on each pixel of the scrambled image.In the simulation experiment and its analysis,the performance of the encrypted image information,such as entropy,shows that the new chaotic image encryption scheme is excellent.Finally,by analyzing the code stream structure of HEVC and the characteristics of video encryption in the entropy coding stage,we choose the syntax elements,that play an important role in image reconstruction and use binary encoder to perform binary arithmetic coding,to encryp.A video encryption arithmetic is proposed,which uses a pseudo-random sequence to randomly choose one of the two types of video encryption algorithms,those are video encryption based on predictive coding and video encryption based on transform quantization process,to encrypt the video sequence.The computer simulation results show that the proposed algorithm has good code stream compatibility and compression ratio invariance,high security,and good anti-exhaustion attack capability. |