Theory analysis of digital hiding technique based on fusion stratagem

In this paper, we will discuss and present a novel digital image hiding technique based on fusion stratagem. It satisfies the requirement of ¿what we saw is not what we get¿. This novel approach to hiding a secret image into the public image using simple Be¿zier curve is proposed on the base of fusion stratagem whose purpose is to hide this secret image. We derive a camouflage image in a fusion way. Under the assumptions that the secret image is a known, the public image signal is an i.i.d.(independent and identically distributed) random sequence from a Gaussian distributed, we show that is this secret image should be extract without the knowledge of strength factor. The extraction is based on maximum-likelihood estimation of the hiding strength of the secret image. The advantage of this method is high efficiency for the secret image hiding and extracting. This method can be used to hide secret image into public images, and the secret image may be an grayscale image ranged from 2 colors to 256 or more color. Experimental results show that the proposed method is feasible.