机构地区: 东莞理工学院计算机学院
出 处: 《计算机工程与设计》 2013年第6期2075-2078,2194,共5页
摘 要: 一些基于标准支持向量机的图像水印技术,使图像水印的效果得到一定改善,但降低了图像水印技术的效率。针对这个问题,用光滑支持向量机取代标准支持向量机,结合图像的局部相关特性来确定图像的最佳嵌入位置和嵌入强度,提出一种基于光滑支持向量机的图像水印技术,并做了仿真实验。实验表明,这种技术与以往基于标准支持向量机的图像水印技术相比,不仅效果更优,而且效率也显著提高。 Using a class of new smoothing functions, the problem of smooth support vector machine (SVM) was studied. A new model of SVM, 3rd-order polynomial smooth support vector machine (3SSVM), was proposed, and its global convergence was established. Numerical experiments were carried out to evaluate 3SSVM, using Newton-Armijo algorithm. The results show that 3SSVM is better than PSSVM and SSVM in the classification performance and computational speed. A better theoretical support is provided for applications of smooth support vector machine.
关 键 词: 光滑支持向量机 标准支持向量机 图像水印 人眼视觉系统 图像局部相关性
领 域: [自动化与计算机技术] [自动化与计算机技术]