摘要
鉴于现有绝大多数图像水印方案将载体图像看作整个场景,并未考虑载体图像自身的结构特征及人们感兴趣的视觉目标区域,本文提出了一种基于感兴趣区(ROI)的小波域图像水印嵌入算法.该算法首先结合人眼视觉感知特性,在小波变换域内利用k-均值聚类提取出感兴趣区域;然后对感兴趣区域进行小波变换,并结合图像局部相关性对小波系数进行自适应的量化调制,将水印信息嵌入到感兴趣区域的低频子带内.仿真实验表明,所提出的水印嵌入算法不仅具有较好的透明性,而且对JPEG压缩、叠加噪声、锐化等常规图像处理操作具有较好的鲁棒性.特别地,该算法能够有效抵御恶意剪切、目标移动、替换背景等攻击,而且可以实现盲检测.
In this paper, a novel robust image watermarking algorithm in region of interest （ROD is proposed,which can solve the problem that the structure characteristics of host image and the interested object region is not taken into account in most existing watermarking scheme. Firstly,the ROI are extracted in DWT domain by using the human visual characteristic and Kmean clustering. Secondly,the discrete wavelet transform （DWT） is performed on the ROI. Finally, the watermark is embedded by adaptively quantizing the low-frequency wavelet coefficients according to local image characteristics. Experimental results show that the proposed algorithm is robust to compression,noise adding,sharpening and various background attacks such as background cropping,ROI moving, background changing.
出处
《小型微型计算机系统》
CSCD
北大核心
2007年第5期920-925,共6页
Journal of Chinese Computer Systems
基金
辽宁省自然科学基金项目(20032100)资助
视觉与听觉信息处理国家重点实验室开放基金项目(0503)资助
大连市科技基金项目(2006J23JH020)资助
"图像处理与图像通信"江苏省重点实验室开放基金项目(ZK205014)资助
江苏省计算机信息处理技术重点实验室开放课题基金项目(KJS0602)资助
关键词
数字水印
感兴趣区
局部相关
digital watermarking
region of interest
local image characteristics
