Image scrambling algorithm based on random shuffling strategy

Conventional scrambling methods based on permuting pixels coordinates can be used to construct robust watermarkings which can endure erasing, cropping and compressing attacks. But most of these methods are used to scramble equilateral image, for the non equilateral image where its width not equal to its height, it is usually expanded into equilateral image or partitioned into several equilateral images, which increases the cost for extra space or operating complexity. Although there are some methods which can scramble non equilateral image, these methods need to construct coordinate shifting path first and the cost to build coordinate shifting path is usually expensive. To address these problems, in this study, we propose a new scrambling algorithm based on random shuffling strategy, which can scramble non equilateral image and has a low cost to build coordinate shifting path. The proposed algorithm has a good one time scrambling performance. It can be used to scramble or recover image in real time and can also resist the JPEG compression attacks. Experiments show the proposed scrambling method validity in scrambling or recovering non equilateral image and robustness in enduring erasing, cropping and JPEG compressing attacks.