论文部分内容阅读
利用遥感数据,综合最大似然法监督分类、多尺度空间分层聚类的部分监督分类方法和主成分方法,分析黄河上游龙羊峡水库库区1987～1999年间土地利用土地覆盖变化。提取专题信息,不同要素采用不同方法;具体分类中,土地利用类型的一级类型耕地、水体及未利用土地类型采用主成分分析和最大似然法监督分类方法;对一级类型草地采用多尺度分层聚类算法的部分监督分类方法。结果表明,草地信息利用SSHC方法提取结果较好,与Bayes分类方法相比,精度提高4.2%,SSHC所获结果数据Kappa系数为0.84,Bayes所获结果数据Kappa系数为0.78。对某专题要素分类,此方法结果较优。
Based on remote sensing data, the method of supervised classification, multi-scale spatial hierarchical clustering based on supervised classification and principal components method, the paper analyzes the land use change of land use in Longyangxia reservoir area from 1987 to 1999 in the upper Yellow River. In the specific classification, the first-class types of cultivated land, water bodies and unutilized land types of land use type are supervised and classified by principal component analysis and maximum likelihood method; the first-class type grassland adopts multi-scale Partially supervised classification method of hierarchical clustering algorithm. The results showed that the grassland information extracted by SSHC method was better, the precision was improved 4.2% compared with Bayes classification method, the Kappa coefficient obtained by SSHC was 0.84, and the Kappa coefficient obtained by Bayes was 0.78. The classification of a thematic feature, this method is better.