Abstract: To lay the watermarking information in the notable feautres areas of video, an algorithm for fuzzy entropy measure(FEM) is proposed. The embedding watermark strength varies with the statistic of the local feature coefficients of the original image, with the modification of these coefficients being self-adaptive. The logo can be detected certainly from the watermarked video sequence by a special Kalman filter without the original video data, and will provide an indication of the copyright existence. Experimental results show that the proposed scheme is effective and turns out to be a reliable authentication tool for the digital Video.

Key words: Video watermarking, copyright protection, fuzzy entropy, chaotic mapping