A Robust Image Hiding Method Based on Sign Embedding and Fuzzy Classification

In order to minimize the distortion of secret image, a novel steganographic method based on sign embedding and fuzzy theory is presented in this paper. By incorporating fuzzy theory with human visual system (HVS), a fuzzy classifier for DCT sub-blocks is first developed to realize the adaptive selection of the embedding strength. The new embedding scheme consists of a reasonable partition of pixel value and a sign embedding operation, which can effectively overcome the grievous distortion problem due to distortions introduced in the secret image by rounding off errors. The experimental results show that the proposed scheme is a robust image hiding approach in DCT domain and the maximum distortion error in pixel value for extracted secret image is plusmn 1. And the cover images can provide good quality after embedding larger amount of data