摘要

One of the greatest advantages of image and video compression employing the wavelet transform is the scalability of both quality and resolution. In this paper, we propose an image and video watermarking technique in wavelet domain designed to survive geometrical and numerical scaling attacks. The watermark extraction is blind, that is, the original un-watermarked image or video is not needed for watermark extraction. Since the binary watermarks, '1' and '0', are embedded by spreading spectrum with reverse polarity, only the polarity of correlation between the watermark pattern and the embedding area is needed for watermark extraction. Besides its survival of resolution scaling, the embedded watermark can be correctly extracted even by brute compression, as well as most of the intentional numerical attacks.

  • 出版日期2010

全文