基于奇异值分解的灰度水印算法  

A gray-level watermarking algorithm based on SVD

作　　者：刘俊清[1] 赵海[1] 黄轶[2] 

机构地区：[1]沈阳工程学院自控系,沈阳110136 [2]华北科技学院电信系,北京101601

出　　处：《信息技术》2010年第11期87-89,共3页Information Technology

摘　　要：提出了一种基于离散小波变换和奇异值分解的灰度水印算法。首先在嵌入前对水印做Arnold置乱处理,然后对置乱后的水印图像及其原图像进行离散小波变换分解,同时保留小波分解后水印的高频系数,并在水印检测过程中作为密钥使用。最后对原图像小波变换后的中低频系数进行奇异值分解(SVD),并把相应的水印嵌入其中。实验结果表明,上述算法能够经受住剪切、噪声、滤波、压缩等处理,具有很强的鲁棒性。This paper proposed a kind of watermark algorithm with gray-level image based on DWT and SVD.Firstly,the watermark is treated with Arnold scrambling transformation before embedding.Then,the scrambled watermark image and original image were decomposed by DWT method.At the same time,the high frequency coefficients of the decomposed watermark were reserved and used as the secret key in the detection process.Finally,the low and middle frequency coefficients of wavelet-transformed original image were treated with SVD,the corresponding watermark was embedded into it.The experimental results show that the above algorithm is robust against cutting,noise,filtering and compression.

关 键 词：灰度水印 离散小波变换 ARNOLD置乱 奇异值分解 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]