Abstract: Because chaotic encryption is a new area of research,most video encryption algorithms are based on low dimensional maps,and their securities need to be improved．Although the capability of encryption of high-dimensional chao-tic system is much more better,the generation process of chaotic sequences are also much more complex．It conflicts with the characteristics of large amount of data and the high real-time requirement of video encryption．Thus the high-dimensional chaotic system is less used in the encryption of video．In light of this,combining with ideas of diffusion of pixels,this dissertation proposed a color video image encryption algorithm based on chaotic Lorenz system．The system follows mechanism of confusion and diffusion proposed by Shannon’s information theory．Any pixel in the image which is decrypted incorrectly will affect the whole decryption of the video．The algorithm has a high security,besides it also has the advantage of fast encryption．The algorithm can satisfy the characteristics of video encryption which have a large quantity of data and demand a high real-time.

Key words: Lorenz chaos,Chaotic attractor,Image encryption,Video encryption