针对遥感图像土地利用一种类别由多种地物组成, 存在难以求取类别光谱特征多元分布模型的问题, 分析了多光谱遥感图像土地利用的光谱特征和区域多中心特征, 提出了一种光谱信息和区域信息基于规则的区域多中心分类方法, 以类别的类内中心集合表征类别模式, 以区域为分类单元, 以区域单元含类别类内中心数和区域单元中属于某种类别的像元占单元总像元的百分比为分类准则; 采用类内中心表征类别模式和基于规则的分类方法, 较好地解决了土地利用类别由多种地物组成、类别模式不满足多元正态分布的问题, 由于类别区域单元多中心特性差异

A convenient multivariate statistical model is in general not available for the multi-spectral feature of land use (LU) class of remote sensing (RS) image, as one LU class is made up of several ground objects. Analyzing the spectral characteristics of LU of multispectral RS imagery, this paper presents a rule-based region multi-center (RMC) method. In the method, the classification cell is pixel region, the classificatory pattern is a set of the classificatory intra-class centers which is confirmed by clustering the training samples, and the classification rules are the type amounts of intra-class center and the percentage of the pixels belonged to the class from the whole region pixels. RMC method can also be used to recognize the individual LU class from multi-spectral RS image. This method deals with the distribution problem by multi-center and based on rule method. There are much difference of the classificatory center amount and the pixel’s percentage among different LU class, so the selection of training area and the determinants of rules are easy. The results of experiment indicate that the LU classification accuracy is increased between 4% and 6% with this method.