|本期目录/Table of Contents|

[1]付 港,李 鸣*.最小二乘支持向量机的自适应数字水印算法[J].武汉工程大学学报,2015,37(06):67-72.[doi:10. 3969/j. issn. 1674-2869. 2015. 06. 014]
 I.Adaptive digital watermarking algorithm based on least squares support vector machine[J].Journal of Wuhan Institute of Technology,2015,37(06):67-72.[doi:10. 3969/j. issn. 1674-2869. 2015. 06. 014]
点击复制

最小二乘支持向量机的自适应数字水印算法(/HTML)
分享到:

《武汉工程大学学报》[ISSN:1674-2869/CN:42-1779/TQ]

卷:
37
期数:
2015年06期
页码:
67-72
栏目:
机电与信息工程
出版日期:
2015-06-30

文章信息/Info

Title:
Adaptive digital watermarking algorithm based on least squares support vector machine
文章编号:
1674-2869(2015)06-0067-05
作者:
付 港李 鸣*
南昌大学机电工程学院,江西 南昌330031
Author(s):
FU GangLI Ming
Nanchang University College of Mechanical and Electrical Engineering, Nanchang 330031,China
关键词:
数字水印最小二乘支持向量机不可感知性鲁棒性粒子群算法
Keywords:
digital watermarking LS-SVM imperceptibility robustness particle swarm algorithm
分类号:
TP309
DOI:
10. 3969/j. issn. 1674-2869. 2015. 06. 014
文献标志码:
A
摘要:
为了提高数字水印的安全性,提出了一种新的基于对比度掩蔽特性的数字水印算法. 根据人眼视觉系统(HVS)的视觉特性,人眼对于亮度信号(Y)的改变较色度信号(U、V)的改变更加敏感,在图像学国际标准CCIR601当中,图像YUV空间中的Y∶U∶V通常会取4∶2∶2. 首先引入对比度敏感函数模型,选择一些对比度函数值最大的子块,并升序排列;接着,对于前一半子块,使用最小二乘支持向量机建立图像邻域像素之间的关系模型;最后,通过动态地调整模型的输出值与中心像素值之间的大小关系,在另一半子块中实现水印的嵌入与提取. 同时,还利用粒子群算法优化了最小二乘支持向量机的超参数. 仿真实验表明,该数字水印算法具有着良好的不可感知性与鲁棒性,对一些常规的攻击操作也有着很强的抵抗能力.
Abstract:
To improve the security of digital watermark, a novel digital watermarking algorithm based on the contrast masking properties was proposed. The human eye to the change of luminance signal (Y)is more sensitive compared with that of the chrominance signals (U、V), according to the visual characteristics of human eye vision system (HVS). In the YUV image space, the values of Y∶U∶V are 4∶2∶2 by the international image standard CCIR601. First, we introduced the contrast sensitivity functions, and chose some sub-images of the maximum contrast sensitivity functions, then ascended them. Second, for the front half of sub-images, the relational model of image neighborhoods pixels was established by the least squares support vector machine (LS-SVM); Finally, the magnitude relation between the output and the central pixel of the model was dynamically adjusted, meanwhile, the watermark was embedded and extracted in the other half of sub-images. The hyperparameters of LS-SVM were also optimized by the particle swarm algorithm. The digital watermarking algorithm has good robustness and imperceptibility, and has a strong resistance to aggressive operation.

参考文献/References:

[1] 邢丹,王雷,乔玉. 一种基于图像统计特性的空域水印算法[J]. 电子技术与软件工程,2015(8):102.XING Dan, WANG Lei, QIAO yu. A airspace watermarking algorithm based on the statistical properties of the image[J]. Electronic Technology & Software Engineering, 2015(8):102.[2] 黄雄华,王宏霞,蒋伟贞,等. 基于BP神经网络和SNR的自适应音频水印算法[J]. 光电子.激光,2012(12):2391-2397.HUANG Xiong-hua, WANG Hong-xia, et al. The adaptive audio watermarking algorithm based on the BP neural networks and SNR[J]. Journal of Optoelectronics. Laser, 2012(12):2391-2397.[3] 王忠,叶雄飞. 遗传算法在数字水印技术中的应用[J]. 武汉工程大学学报,2008,30(1):95-97.WANG Zong,YE Xiong-fei. The application of genetic algorithm in the digital watermarking technology[J]. Journal of Wuhan Institute of Technology,2008,30(1):95-97.[4] HU R, CHEN F, YU H M. Incorporating Watson’s perceptual model into patchwork watermarking for digital images[C]//17th IEEE International Conference on Image Processing, 2010:3705-3708.[5] 张力,邓亚航,饶小李. 颜色特征模型在静态车辆检测中的应用[J]. 武汉工程大学学报,2015(1):73-78.ZHANG Li, DENG Ya-hang, RAO Xiao-li.Color feature model in application of static vehicle detection[J]. Journal of Wuhan Institute of Technology, 2015(1):73-78.[6] STEVE R G. Support vector machines for classification andregression[R].Technical report, Southampton University of Southamp-ton, Faculty of Engineering, Science and Mathematics Scod of Electronics and Computer Science, 2003:1-28.[7] 李松,顾巧论,高铁杠. 信息嵌入位置自适应选择的鲁棒水印算法[J]. 计算机应用,2013(2):438-440,446.LI Song, GU Qiao-lun, GAO Tie-gang. Robust watermark algorithm for adaptive choice of information embedding position.[J] Journal of Computer Applications, 2013(2):438-440,446.[8] 何娟芝.基于YUV颜色空间的边缘检测算法研究[D].杭州:杭州电子科技大学,2013.HE Juan-zhi. A study of the edge detection algorithm based on the YUV color space[D].Hangzhou:hangzhou dianzi university, 2013.[9] 陈军,胡福乔.YUV空间的彩色图像HDR合成算法[J]. 计算机工程,2012(9):214~216,233.CHEN Jun, HU Fu-qiao. A HDR synthesis algorithm of the color image in the YUV space[J]. Computer Engineering, 2012(9):214~216,233.[10] 吴明巧. 数字图像信息隐写与隐写分析技术研究[D]. 长沙:国防科学技术大学,2007.WU Ming-qiao. The research on steganorgraphy and steganalysis of digital image[D]. Changsha:National university of defense technology, 2007.[11] 倪顾伟. 基于神经网络的数字水印算法的研究与实现[D]. 南京:南京理工大学,2012.NI Gu-wei, The study and implementation of the digital watermarking algorithm based on the neural network[D]. Nanjing: Nanjing university of science and technology, 2012.

相似文献/References:

[1]王忠,叶雄飞.遗传算法在数字水印技术中的应用[J].武汉工程大学学报,2008,(01):95.
 WANG Zhong,YE Xiong fei.The application of genetic algorithm in digital watermark[J].Journal of Wuhan Institute of Technology,2008,(06):95.
[2]付 港,李 鸣*.纹理掩蔽特性在水印技术中的应用[J].武汉工程大学学报,2015,37(07):55.[doi:10. 3969/j. issn. 1674-2869. 2015. 07. 012]
 *.Application of texture masking characteristics in watermark technology[J].Journal of Wuhan Institute of Technology,2015,37(06):55.[doi:10. 3969/j. issn. 1674-2869. 2015. 07. 012]
[3]王华杰,郑来芳.电动汽车电池荷电状态估算[J].武汉工程大学学报,2015,37(10):51.[doi:10. 3969/j. issn. 1674-2869. 2015. 10. 010]
 .Evaluating charge state of electric vehicle battery[J].Journal of Wuhan Institute of Technology,2015,37(06):51.[doi:10. 3969/j. issn. 1674-2869. 2015. 10. 010]
[4]张正风.基于LS-SVM苯乳酸发酵过程的建模[J].武汉工程大学学报,2016,38(4):333.[doi:10. 3969/j. issn. 1674?2869. 2016. 04. 005]
 ZHANG Zhengfeng.Modeling of Phenyllactic Acid Fermentation Process Based on Least Square Support Vector Machine[J].Journal of Wuhan Institute of Technology,2016,38(06):333.[doi:10. 3969/j. issn. 1674?2869. 2016. 04. 005]

备注/Memo

备注/Memo:
收稿日期:2015-05-14作者简介:付港(1988-),男,湖北汉川人,硕士研究生.研究方向:精密仪器与机械.* 通信联系人.
更新日期/Last Update: 2015-08-24