Mosaic secret-fragment-visible data hiding for secure image transmission based on two-step energy matching

Recently, a new information hiding technology was proposed to hide a secret image into secret-fragment-visible mosaic image for secure image transmission. However, the previous method pays less attention to the block matching in the transform, and may obtain a low quality mosaic transform in some cases. In this paper, we propose to improve the secret-fragment-visible data hiding based the two-step energy function, which returns a weighted value of standard deviation and total variation of block. We match all the blocks according to the energy function and take the zero-mean normalized cross correlation (ZNCC) as the similarity measurement to optimize the matching. The interpolation optimization algorithm is used to stimulate the total ZNCC changes along with the parameter selection, and helps to solve the optimization quickly. Since more block's statistical properties are employed, the image blocks can be matched accurately, and a low distortion is introduced in the transform. Experimental results show that the proposed method can obtain a better visual quality of secret image after extraction and also requires less bits to encode the residuals.