Bi-level Fragmented Scrambling for Image Secrecy

Image position scrambling is the pre-processing stage in image cryptography and image steganography. Arnold’s cat map is one of the most commonly used methods in pixel location scrambling. Since the method is performed on the all pixels of the whole image one by one, it takes time so much that limits its application. By the way, due to the image dimensions are obvious; the image can be recovered simply at most with T (period) iterations, so it needs an extra stage of encryption. This paper has tried to reduce the pixel location scrambling time and to increase its secrecy and security. In this paper a new method is proposed which apply Arnold’s transformation in two different levels. So it is called Bi-level Fragmented Scrambler, BFS. BFS fragments the image to a predefined number in each level. These predefined values can be considered as a key for the image scrambler. The method is simulated to find the computation time and a scrambling degree, to compare BFS in different conditions (i.e. for different image size and number of fragments) and with the previous method. Simulation results show that BFS can increase security in the scrambling step (so no extra encryption stage will be needed), while reduces the process time effectively.

کلیدواژه ها:

Image processing ، Scrambling ، Encryption ، Pixel location scrambler ، Arnold matrix