【摘要】 该文对应用高程、坡向及土壤等辅助信息提高分类精度进行了研究 .在GIS支持下 ,运用GIS的空间分析技术综合研究高程、坡向等辅助信息与植被类型的内在关系 ,运用模糊数学的基本原理将这种内在关系量化 ,形成可以反映植被与辅助信息间规律的模糊矩阵 ,综合应用所得到的模糊矩阵通过后分类法来修正无监、有监分类得到的初分类图像 .该研究以贺兰山中段汝箕沟一带为研究对象 ,成功地应用该方法得到了研究区的植被分类图像 .研究表明 ,三维及相关辅助信息可以有效地提高遥感图像的分类精度更多还原

【Abstract】 This paper found the relations between the vegetation and ancillary data such as elevation, aspect and soil types through the spatial analysis. The authors quantified the relations through the form of fuzzy matrix, using basic principles of fuzzy math. With the fuzzy matrix, the paper developed a post classification method to correct the “error areas” in the Landsat TM classification of supervised method. Using the post classification method, the authors successfully acquired the vegetation classification images in the test area Helan Mountain, which lies in North China. The paper shows that the ancillary data can improve the vegetation classification accuracy effectively.更多还原