Image Authentication Using Stochastic Diffusion

This paper considers an approach to encrypted information hiding based on Stochastic Diffusion for encrypting digital images coupled with the application of a Least Significant Bit (LSB) method for information embedding. After providing a brief summary of various information hiding methods based on spatial and transform domain techniques, two new methods are introduced. The first of these considers a binary image watermarking algorithm for hiding an image in a single host image which is based on binarization of the encrypted data. The second method extends this approach to solving the problem of 24-bit image hiding in three host images which generates a near perfect reconstruction after decryption. Both methods make use of a 'hidden code' technique to randomize the order of the embedded bits and the location (in the image plane) of the LSBs which make the embedded information more robust to attack. Details of the algorithms developed are provided and examples are given, which have application in the field of covert cryptography and the authentication of full colour images for copyright protection and Data Rights Management.