基于Logistic混沌映射和IWT-SVD量化的盲鲁棒水印算法  

Blind robust watermarking algorithm based on Logistic chaotic mapping and IWT-SVD quantization

作　　者：蒋晓丹 范航宇 陆哲明[2] 

机构地区：[1]衢州学院电气与信息工程学院,浙江衢州324000 [2]浙江大学航天航空学院,浙江杭州310027

出　　处：《传感器与微系统》2018年第2期131-135,共5页Transducer and Microsystem Technologies

基　　金：浙江省科技计划公益技术研究项目(2016C31097);信息安全国家重点实验室开放课题项目(2017MS16);衢州市科技计划资助项目(2015Y005)

摘　　要：为了提高数字水印的鲁棒性以及水印嵌入位置的保密性,提出了一种基于Logistic混沌映射和整数小流变换—奇异值分解(IWT-SVD)量化的盲鲁棒水印算法:对二值水印图像进行Arnold置乱,对原始载体图像进行8×8分块,利用生成的Logistic混沌序列选择嵌入水印的图像块,充分保证了水印嵌入位置的保密性;对选中的图像块进行二级IWT分解,对得到的二级低频子带系数进行SVD;将置乱后的水印信息嵌入。算法不需要原始载体图像,实现了水印的盲提取,更具实用性。仿真实验结果表明:算法对常规图像处理操作、压缩攻击和剪切攻击均具有较强的稳健性。A blind robust watermarking algorithm based on Logistic chaotic mapping and integer wavelet transform（ IWT）-singular value decomposition（ SVD） quantization is proposed in order to enhance robustness and position security of the embedded watermark. Binary watermarking image scrambling is carried out using Arnold transform.The original host image is divided into 8 × 8 blocks,and the image blocks to be embedded with watermark bits are selected by using Logistic chaotic sequence,which can enhance the position security of the embedded watermark.The selected image blocks are transformed by IWT,and the coefficients of two-level low-frequency subbands are decomposed with SVD. The scrambling watermark information is embedded into those coefficients. During the watermark extraction,the algorithm does not require original host image,which realizes blind extraction of watermark,which makes the algorithm more practical. Simulation results show that the proposed algorithm is robust in conventional image processing operations,compression attacks and shear attacks.

关 键 词：鲁棒水印 混沌映射 整数小波变换 奇异值分解 量化 盲提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]