Image authentication and recovery using wavelet-based multipurpose watermarking

In this paper, a multipurpose image watermarking method based on the wavelet transform is proposed for content authentication and recovery of the tampered regions. In the proposed method, the original image is first divided into nonoverlapping blocks and each block is transformed into the wavelet domain. The image features are subsequently extracted from the lowest-frequency coefficients of each block as the first embedded watermark. Next, the whole image is decomposed into the two-level wavelet transform and the orientation adjustment is calculated based on the wavelet coefficients in the middle-frequency subbands for image authentication. In addition, a logo watermark is embedded into the given middle-frequency subbands. Therefore, the copyright and feature of the image can be held on after two different watermarks are embedded. Experimental results show that the proposed method is efficiency for accurate detection of tampered regions and recovery of estimated image.