基于多特征组合的彩色遥感图像分类研究

为了解决利用单一特征对彩色遥感图像进行分类效果不理想、普适性不强等问题，提出了一种基于颜色和纹理特征组合的支持向量机彩色遥感图像分类方法。该方法尝试将彩色遥感图像的颜色信息和纹理信息相结合作为支持向量机算法分类的特征向量，据此对遥感影像进行分类，并进行了实验验证。结果表明，颜色和纹理特征组合的支持向量机分类方法能够取得较高的分类精度，其分类效果优于传统的单一颜色或纹理特征分类，是一种有效的彩色遥感图像分类方法。

Classification of color remote sensing images based on multi-feature combination

In order to improve the classification results and solve the universality in color sensing image classification using unique feature, a new support vector machine (SVM) color remote sensing image classification algorithm based on color feature and texture feature combination was proposed. The method used the combination of the color information and the texture information of color remote sensing image as the eigenvectors of SVM algorithm. The results show that the method can achieve higher precision compared with the traditional method using unique feature or texture feature. The method is effective to classify the remote sensing image.