摘要
基于对象的视频水印在几何变换的攻击下将丧失同步信息.采用图像规范化的方法,比较水印嵌入和检测时的规范化信息来达到同步检测,并进一步校正了由于插值误差产生的波动点.同时为了抵抗MPEG- 4压缩,通过分析水印嵌入量、鲁棒性和视频质量三者的关系,使用编码阶段量化后的四个SA- DCT系数来嵌入一个水印分量.算法可方便地并入到MPEG- 4基于对象的编码框架中.实验结果表明本方案具有失真小、鲁棒、简单易行及可操作性强等优点.
Geometric transformations often disable the detection of object-based video watermarking systems because of lacking synchronization information. Image normalization was used to achieve synchronization by comparing the normalization info before watermark embedding/detection. Then the fluctuating points caused by interpolation when normalizing was adjusted. In order to be robust to the MPEG-4 compression, this paper embeds the watermark in the quantized SA-DCT coefficients. Especially, four such coefficients are selected to embed one watermark bit according to the relationships among watermark capacity, robustness and video quality. The algorithm can be easily incorporated into the MPEG-4 object-based coding framework. Experiments show that the scheme was low-distortion, robust, simple and easy to use.
出处
《小型微型计算机系统》
CSCD
北大核心
2005年第5期803-806,共4页
Journal of Chinese Computer Systems
基金
国家自然科学基金会重点项目研究计划 (60 13 3 0 2 0 )资助
关键词
视频对象
水印
图像规范化
几何变换
video objects
watermarking
image normalization
geometric transformation