Multi-Image Hiding Technology With Fractional Mellin Transform

Data hiding technology, which embeds secret information into host data to transmit, such as image, voice, video, document, and etc. has drawn more and more attention and become one of the hot research topics in the field of information security. In the field of image encryption, the nonlinear fractional Mellin transform performs well. Therefore, it is of important significance to apply fractional Mellin transform to image hiding. Two multi-image hiding algorithms with fractional Mellin transform are proposed in this dissertation. The algorithms with considerable embedding capacity could increase the hiding information. The main research results are as follows:A multi-image hiding scheme based on fractional Mellin transform is proposed.Multiple secret images with M ?N pixels are transformed by the discrete cosine transform(DCT). The transformed images are truncated with the coefficient M ?N n and combined into one image by using the method of spectrum truncation and splicing.The splicing image and the host one are divided into several annular domains and then transformed by the fractional Mellin transform of different orders, respectively. The transformed result of the splicing image is embedded into the coefficients of the fractional Mellin transform on the host image with a Bezier curve randomly. The fused image can be obtained with the inverse fractional Mellin transform of the same orders on the host image, which contains the information of the secret image. Several groups of images are used in the simulation, and the results show that the proposed method is feasible with certain robustness to resist noise attack.A multi-image hiding method based on fractional Mellin transform and DCT coefficient modification is introduced. Multiple secret images are transformed by fractional Mellin transform of different orders. And the outputs are encrypted into one cipher-text by an iteration process. Then the encrypted image is embedded into the DCT coefficients of the host image. The hidden image could be obtained from the modified coefficients of the DCT by taking inverse DCT, which is visually identical to the host image and contains the information of the secret image. The quality of the hidden image decreases with the increasing of the embedding strength. Simulation results demonstrate that the proposed method is sensitive to keys with large key space and good robustness.