基于GF-2遥感影像的甘东南农村建筑物结构分类方法  

Classification Method for Rural Building Structures in Southeast Gansu Province based on Remote Sensing Images

作　　者：孙钦瑶 钟秀梅[1,2] 马金莲 王妍 许晓威 吴松翰 王谦[1,2,3] SUN Qinyao;ZHONG Xiumei;MA Jinlian;WANG Yan;XU Xiaowei;WU Songhan;WANG Qian(Lanzhou Institute of Seismology,CEA,Lanzhou 730000,China;Key Laboratory of Loess Earthquake Engineering of CEA&Gansu Province,Lanzhou 730000,China;Research Center for Conservation of Cultural Relics of Dunhuang,Dunhuang 736200,China)

机构地区：[1]中国地震局兰州地震研究所,甘肃兰州730000 [2]中国地震局(甘肃省)黄土地震工程重点实验室,甘肃兰州730000 [3]甘肃省敦煌文物保护研究中心,甘肃敦煌736200

出　　处：《遥感技术与应用》2025年第1期192-201,共10页Remote Sensing Technology and Application

基　　金：中央级科研院所基本科研业务费(2021IESLZ03);甘肃省敦煌文物保护研究中心开放课题(GDW2021YB07);中国地震局地震工程与工程振动重点实验室重点专项(2020EEEVL0409);甘肃省地震局创新团队计划(2021TD⁃01⁃01)。

摘　　要：农村建筑物作为地震灾害中最重要的承灾对象,对其类型、分布等信息的快速获取在抗震减灾等方面具有重要意义。基于GF-2高分辨率遥感数据,利用ESP(Estimation of Scale Parameter)算法和SeaTH(Seperability and Thresholds)算法分别确定影像最佳分割尺度及构建最优特征学习空间,选用决策树分类法和随机森林机器学习分类法,分别对2021年5月初甘肃省襄南镇的农村建筑物结构进行提取分类,并使用无人机航测和现场调查统计数据进行分类结果的准确度检验和修正。结果表明:①两种方法都能较好地识别空间分布均匀、面积大、颜色鲜明的砖混建筑物,但对于分布杂乱且相对集中、颜色灰暗、面积小的土木(砖木)建筑物难以有效识别出其边界轮廓并准确分类。②两种方法对研究区建筑物分类的精度分别是82.42%、86.82%,且基于随机森林的方法在提取建筑物信息时出现的错分漏分现象较少,因此,随机森林方法进行农村建筑物分类更适用。The rural buildings are the most important disaster recipient in the earthquake disaster,it has signifi⁃cant meaning in the fields of earthquake resistance and hazardous reduction to obtain the information like the type and distribution of it.Based on GF-2 high-resolution remote sensing data,the ESP(Estimation of Scale Parameter)algorithm and Seath(Seperability and thresholds)algorithm are used to determine the optimal im⁃age segmentation scale and construct the optimal feature learning space.The decision tree classification method and random forest machine learning classification method were chosen to extract and classify rural building struc⁃tures in Xiangnan Town,Gansu Province,in early May 2021.Unmanned aerial survey and on-site investigation data were used to verify and refine the accuracy of the classification results.The results show that:①Both meth⁃ods can better identify brick-concrete buildings with uniform spatial distribution,large area and bright color,but for civil buildings with chaotic distribution and relatively concentrated,gray color and small area(brick-wood buildings)are difficult to effectively identify their boundary contours and accurately classify them.②The accu⁃racy and Kappa coefficient of the two methods for building classification in the study area are 84.42%,86.82%and 0.7015,0.7591,respectively,and the random forest-based method has less misclassification and missing phenomenon when extracting building information.Therefore,the random forest method is more suitable for ru⁃ral building classification.

关 键 词：遥感影像 建筑物结构分类 Seath算法 决策树 随机森林 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]