A hybrid DWT-SVD copyright protection scheme based on k-means clustering and visual cryptography

In this paper, a novel copyright protection scheme that combines the discrete wavelet transform (DWT) and the singular value decomposition (SVD) is proposed. Instead of modifying the original host image to conceal a secret image, the proposed scheme first extracts the image features from the host image by applying the DWT and the SVD. The extracted features are then classified into two clusters by employing the k-means clustering technique, and a master share is generated using the clustering result. Finally, the master share is used together with a secret image to construct an ownership share according to a two-out-of-two visual cryptography (VC) technique. When rightful ownership needs to be determined, the secret image for ownership identification can be revealed by stacking the master share and the ownership share. Experimental results demonstrate that the proposed scheme can effectively resist several common attacks. In addition, the revealed watermark can be easily recognized by human eyes, even if the host image has undergone severe attacks.