基于SVM的图像特征样本集构造技术研究

Research on Sample Sets Construction Technology Based on SVM Image Characteristics

针对图像的过滤、图像水印的检测等方面的研究成为一个热点,统计学习理论成为解决此类问题的一条捷径。介绍了SVM基础理论以及特征向量的意义,探讨了图像特征向量的一般选择方法以及基于粗集的图像特征选择方法,阐述了训练样本集的构造。

The research on the image filter, the detection of the image watermarking has become a hot point, and the mathematical theory of statistics has become a shortcut for solving such kind of problems. This paper introduces the fundamental theory of SVM and the significance of the eigenvector, probes into some common methods for selecting the image eigenvectors and the methods for selecting the image characteristics based on the rough set, and expounds the structure of the training sample set.