基于SVM的墙地砖颜色自动分类

支持向量机(SVM)是一种采用结构风险最小化原则的新的机器学习方法，具有完备的理论基础。该文首次把支持向量机技术应用于墙地砖的自动分类，首先通过对墙地砖图像的RGB通道进行小波分解，由于不同通道的相关性，故提取其协变信号作为特征集，再构建二叉树形式的决策树来实现SVM多类分类，然后对墙地砖进行了颜色分类实验，并与knn分类结果对比，实验结果证明SVM分类器具有更高的分类准确率。

Color Grading of Tiles Based on SVM

Support vector machine(SVM) is a new machine learning algorithm based on structural risk minimization inductive principle. In this paper, the SVM technology is first applied to ceramic tile's automatic classification. First, the RGB channels of tile image are decomposed by wavelet analysis. Then the combinated characteristic set is acquired because of the channels relativity. The classification capability of SVM is extended from two-class classifier to multiple-class classifier by self-organized sequential classification processing. And color grading of tiles is made. The result of experiment shows the SVM classifier has higher correctness than that of the knn classifier.