We have invented a cipher system algorithm based on Chaotic Neural Network (CNN) for video and image encryption. We integrate this innovation with MPEG-2 video compression system to produce an integrated system model able to encrypt and compress video signal. This integrated system model satisfies high security, bit rate and video quality control and it is suitable for secure video transmission applications like medical and military applications. Two main contributions and advantages are discussed in this invention. The first one is related to chaos properties and its effect on CNN cipher algorithm. Results show that very small key modification in the receiver or transmitter side will lead totally to unclear video scene with low Peak Signal to Noise Ratio (PSNR) video quality value. The second advantage is related to bit rate and video quality control of the whole integrated system model depending on bandwidth and communication channel conditions.