基于3D-DCT的RGB图像双水印设计与实现

传统基于量化的水印方法抗JPEG压缩性能较好,但是抗噪声的性能较差,而基于关系的水印方法对噪声等攻击具有较好的性能。因此,为了进一步提高嵌入水印信号的鲁棒性能,提出一种新的基于三维离散余弦变换的RGB图像双水印方案。首先,为了提高嵌入二值水印图像的安全性,设计了一种基于混沌的加密算法对选择的水印信号进行加密处理;其次对载体图像进行互不重叠的分块并随机选择分块进行三维离散余弦变换;最后分别采用关系和量化方法嵌入加密后的水印信号。提取水印时,分别采用关系和量化方法提取出两个水印信号并选择与原始水印相似性较好的信号作为最终提取的水印信号。开发了一个数字水印仿真系统对本文算法进行测试,结果表明本文算法具有较好的不可感知性和抵抗多种攻击的能力。将本文算法与同类算法相比较,对大多数的攻击本文算法具有更优越的性能。

Design and Implementation of RGB Image Dual Watermarking Based on 3D-DCT

The quantization-based watermarking method has better performance for JPEG compression, but the performance against noise is poor, and the relation-based watermarking method has better performance against noise and other attacks. Therefore, in order to enhance the robustness of embedded watermarking signal, a new color image dual watermarking scheme based on three-dimensional discrete cosine transform (3D-DCT) is proposed. First, in order to improve the security of the embedded watermarking, the binary watermarking image is preprocessed by chaotic-based encryption algorithm. Second, the cover color image is divided into non-overlapping blocks and performs 3D-DCT on randomly selected block. Finally, the preprocessed binary watermarking image is embedded by relationships embedding method and quantification embedding method, respectively. When extracting the watermarking, the watermarking signals are extracted by using relation extraction method and quantization extraction method respectively and the extracted signal with better performance is selected as the final watermarking image. A digital watermark simulation system was developed and the experimental results show that the proposed algorithm has better imperceptibility and resistance to a variety of attacks. Compared with similar algorithms, it has a better performance and can be used to protect the copyright of digital products.