A noise resilient scrambling scheme for noisy transmission channel

Image scrambling is an efficient and fundamental scheme used in pre/post-processing of image encryption, image watermarking and information hiding schemes to protect digital images. In this paper, a new image scrambling algorithm based on chaotic systems is proposed. The proposed scheme has multiple internal levels of operation. At each level, the plain-image is first decomposed into same sized blocks which are scrambled using Arnold transform. To strengthen the scheme, the parameters of scrambling are randomly generated through another 2D chaotic system, which makes scrambling key-dependent. Experimental analysis shows that the algorithm is efficient and provides good scrambling effect. Furthermore, it exhibits large key space, high security and PSNR analysis shows that it is resilient towards the transmission channel noise.