基于遥感数据的土地资源分类方法比较研究  

Comparative Study on Land Resource Classification Methods Based on Remote Sensing Data

作　　者：隋瑜 徐泮林[1] 谷彦斐 SUI Yu;XU Panlin;GU Yanfei(College of Geomatics,Shandong University of Science and Technology,Qingdao 266590,China)

机构地区：[1]山东科技大学测绘科学与工程学院,山东青岛266590

出　　处：《测绘与空间地理信息》2021年第2期107-110,共4页Geomatics & Spatial Information Technology

摘　　要：以安徽省安庆市怀宁县为研究对象,通过对该地区的遥感数据处理,采用K-Means、最大似然、最小距离、基于专家知识的决策树4种分类方法进行土地资源分类,对分类结果图进行比较,然后进行分类后处理和精度评价,对分类精度进行比较,比较各方法的准确性和适用性。研究结果表明,针对该研究区,决策树分类结果精度更高。This paper takes Huaining County,Anqing City,Anhui Province as the research object,and uses the four classification methods of K-Means,maximum likelihood,minimum distance and decision tree based on expert knowledge to classify land resources by remote sensing data processing in the area.Compare the classification result graphs,then perform post-classification processing and accuracy evaluation,compare the classification accuracy,and compare the accuracy and applicability of each method.The research results show that the decision tree classification results are more accurate for the study area.

关 键 词：遥感数据 土地资源分类 精度评价 分类方法比较 

分 类 号：P237[天文地球—摄影测量与遥感]