能量差值信息隐藏(differential energy watermarking,DEW)算法通过有选择地舍弃视频图像中某些区域的部分高频DCT系数,形成能量差异而达到信息嵌入的目的.为此,利用视频序列时间域的强相关性建立针对性统计估计模型,侦测DEW算法所造成的载体信息统计特性变化,并结合K-S统计假设检验方法设计一种视频信息隐藏分析算法.结果表明,此算法能够有效地检测压缩视频码流中是否有DEW算法嵌入的特定信息.
A multiple watermarking algorithm is presented according to the multiple accessing technique of the code division multiple access （CDMA） system. Multiple watermarks are embedded into digital images in the wavelet transform domain. Each of the watermarks is embedded and extracted independently without impacts to each other. Multiple watermarks are convolution encoded and block interleaved, and the orthogonal Gold sequences are used to spread spectrum of the copyright messages. CDMA encoded water-mark messages are embedded into the wavelet sub-bands excluding the wavelet HH1 sub-bands. The embedment amplitude is decided by Watson＇ s perceptual model of wavelet transform domain, and the embedmeut position in the selected wavelet sub-bands is decided randomly by a pseudo-random noise （PN） sequence. As a blind watermm＇king algorithm, watermarks are extracted without original image. The watermarking capacity of proposed algorithm is also discussed. When two watermarks are embedded in an image at the same time, the capacity is larger than the capacity when a single watermark is embedded, and is smaller than the sum of the capacity of two separately embedded watermarks. Experimental results show that the proposed algorithm improves the detection bits error rate （BER） observably, and the multiple watermarks have a preferable robustness and invisibility.
