应用P-Fibonacci加密的模糊自适应水印算法  

A Novel Fuzzy Adaptive Watermarking Algorithm Based on P-Fibonacci Encryption

作　　者：冯祥斌[1] 陈永红[1] 

机构地区：[1]华侨大学计算机科学与技术学院,福建厦门361021

出　　处：《华侨大学学报（自然科学版）》2014年第3期287-292,共6页Journal of Huaqiao University(Natural Science)

基　　金：福建省自然科学基金资助项目(2011J05151);华侨大学高层次人才科研启动项目(10Y0199);教育部留学回国人员科研启动基金资助项目(2009年度)

摘　　要：结合P-Fibonacci加密算法和模糊归类,提出一种自适应水印新算法.利用P-Fibonacci加密具有良好的均匀性及安全性,将水印图像的像素位置打乱,消除了二维数字水印图像的像素空间相关性,使嵌入水印后的块效应降低.对原始载体图像进行分块,结合人类视觉系统的纹理和照度掩蔽特性,得到纹理模糊函数和各图像子块的归类结果,并自适应地确定水印的嵌入强度.对载体图像的各子块进行离散小波变换(DWT),将经P-Fibonacci加密算法加密后的水印信息重复嵌入各子块DWT域的低频部分.实验结果表明:该算法抗JPEG压缩、高斯噪声、滤波等常见的信号攻击和几何攻击具有很好的鲁棒性.A new adaptive watermarking algorithm based on P-Fibonacci encryption and fuzzy classified is proposed. The P-Fibonacci encryption has a good uniformity and safety, disrupting the watermark image＇s pixel position by P-Fibonacci encryption algorithm, which can eliminate the spatial correlation of the two-dimensional digital watermark image pixels, and ultimately, reducing the blockiness when embedding watermark information. Simultaneously, blocked the original carrier image, and according to the texture and illumination masking characteristics of the human visual system, obtained a fuzzy function which based on the texture characteristics of the original image, establishing the texture fuzzy function and the classification results of each sub-block, according to the results, determined the embedding strength adaptively. Then discrete wavelet transform （DWT） is performed on the sub-images, and the encrypted watermarks were embedded in the low frequency of the DWT domain repeatedly. The experimental results show that the presented algorithm has a good robustness against common signal processing attacks such as JPEG compression, Gaussian noise, filtering and geo- metric attack such as cropping.

关 键 词：P-Fibonacci变换 FIBONACCI P-CODE 位平面 数字水印 模糊归类 自适应 离散小波变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]