基于BEMD的数字图像水印算法

数字水印技术作为信息隐藏在多媒体领域得到了广泛应用,它的主要思想是将信息的秘密嵌入进图像视频等数字信息产品中,使信息版权得到有效地保护。经验模态分解是一种全新的多尺度分析方法,在非线性非平稳信号分析方面有良好的性能。文章实现了将二维经验模态分解方法应用于数字水印的嵌入与提取,对水印图像添加高斯白噪声和JPEG压缩图像进行攻击并进行水印再提取,验证了算法具有一定的不可见性和较强的鲁棒性。

Digital watermarking technology as information hiding has been widely used in the field of multimedia. Its main idea is to embed the secret of information into digital information products such as image and video, so that the information copyright can be effectively protected. Empirical mode decomposition is a new multi-scale analysis method, which has good performance in nonlinear non-stationary signal analysis. In this paper, the two-dimensional empirical mode decomposition method is applied to the embedding and extraction of digital watermark. White Gaussian Noise and compressed JPEG image are attacked and the watermark image are re-extracted, which proves that the algorithm has certain invisibility and strong robustness.