基于决策树模型的采煤沉陷湿地信息提取  

Remote Sensing Extraction of Aquatic Vegetation Information in Coal Mining Subsidence Wetland Based on the Decision Tree Model

作　　者：郭辉[1,2,3] 韩修壮 GUO Hui;HAN Xiuzhuang(School of Geomatics,Anhui University of Science and Technology,Huainan Anhui 232001,China;Key Laboratory of Aviation-Aerospace-Ground Cooperative Monitoring and Early Warning of Coal Mining Induced Disasters of Anhui Higher Education Institutes,Anhui University of Science and Technology,Huainan Anhui 232001,China;Coal Industry Engineering Research Center of Mining Area Environmental and Disaster Cooperative Monitoring,Anhui University of Science and Technology,Huainan Anhui 232001,China)

机构地区：[1]安徽理工大学空间信息与测绘工程学院,安徽淮南232001 [2]安徽理工大学矿区环境与灾害协同监测煤炭行业工程研究中心,安徽淮南232001 [3]安徽理工大学矿山采动灾害空天地协同监测与预警安徽普通高校重点实验室,安徽淮南232001

出　　处：《安徽理工大学学报（自然科学版）》2022年第4期64-70,共7页Journal of Anhui University of Science and Technology:Natural Science

基　　金：安徽理工大学引进人才科研基金(13200002);矿山空间信息技术国家测绘局重点实验室开放项目(KLM201801)。

摘　　要：为探求高潜水位沉陷湿地水生植被信息遥感提取的有效方法,以潘集矿区采煤沉陷湿地为研究对象,构建水生植被信息提取的决策树模型。选用研究区2021a 9月11日的Sentinel-2影像数据验证模型的有效性,结果显示决策树模型的总体分类精度为83.1%,Kappa系数为0.78,高于支持向量机、神经网络和分类与回归树法方法的分类精度。在只针对浮叶植被和挺水植被进行提取的情况下,决策树模型的总体分类精度和Kappa系数可以达到91.3%和0.81,而其他3种分类方法没有明显的提高,说明构建的决策树模型在水生植被的提取上更具优势。决策树模型普适性检验的结果表明,该模型可应用于矿区其他沉陷湿地水生植被的提取,为矿区湿地植被资源监测提供技术参考。In order to explore an effective method for extracting aquatic vegetation information from high groundwater level subsidence wetland by remote sensing,a decision tree model for extracting aquatic vegetation information was established based on the coal mining subsidence wetland in Panji mining area.The sentinel-2 image data of the study area on September 11,2021 was selected to verify the effectiveness of the model.The results showed that the overall classification accuracy of the decision tree model and the kappa coefficient were 83.1%and 0.78 respectively,higher than the classification accuracy of support vector machine,neural network and classification and regression tree methods.In the case of extracting only floating leaf vegetation and emergent vegetation,the overall classification accuracy and kappa coefficient of the decision tree model would reach 91.3%and 0.81,while the other three classification methods were not significantly improved,indicating that the decision tree model constructed in the research has more advantages in the extraction of aquatic vegetation.The results of the universality test of the decision tree model showed that the model was able to be applied to the extraction of aquatic vegetation in other subsidence wetlands in the mining area,providing a technical reference for the monitoring of wetland vegetation resources in the mining area.

关 键 词：采煤沉陷湿地 Sentinel-2 决策树模型 信息提取 

分 类 号：P237[天文地球—摄影测量与遥感]