文章摘要

文章信息/Info

Title:: Study on the object-based classification of low-altitude UAV remote sensing image of the Lushan earthquake based on the SEaTH algorithm

作者:: 王之¹; 2; 刘超¹; 2; 刘秀菊³; 鲁恒¹; 2; 蔡诗响¹; 2; 杨正丽¹; 2; (1.四川大学水力学与山区河流开发保护国家重点实验室,四川成都 610065; 2.四川大学水利水电学院,四川成都 610065; 3.成都市规划信息技术中心,四川成都 610042)

Author(s):: WANG Zhi¹; 2; LIU Chao¹; 2; LIU Xiuju³; LU Heng¹; 2; CAI Shixiang¹; 2; YANG Zhengli¹; 2; (1. State Key Laboratory of Hydraulics and Mountain River Engineering,Sichuan University,Chengdu 610065,Sichuan,China)(2. College of Hydraulic and Hydroelectric Engineering,Sichuan University,Chengdu 610065,Sichuan,China)(3. Chengdu Planning Information Technology Center,Chengdu 610042,Sichuan,China)

Keywords:: object-oriented classification; multiresolution segmentation; the SEaTH algorithm; UAV images

摘要:: 在地震这种重大自然灾害面前,快速有效地从遥感影像中提取震区土地利用信息,在灾情评估及灾后重建中发挥着重要作用。选取四川省芦山地震灾区无人机影像为数据源,运用面向对象的影像分析方法,首先研究了多尺度分割中参数的选择,获取了研究区最优分割参数; 然后考虑了各个“影像对象”的数字化特征值,利用改进的SEaTH算法进行特征值优化处理; 最后采用了隶属度信息提取方法,获得了芦山地震灾区无人机低空遥感影像分类图,并进行了分类精度评估,结果表明:研究区影像的分类总精度为87.5%,kappa系数为0.835。

Abstract:: Before an earthquake that is believed to be a major natural disaster,how to quickly and efficiently extract the area of land use information from remote sensing images,plays a role in the evaluation of the disaster and post-disaster reconstruction. In this study,we select UAV images of the Lushan earthquake stricken areas in Sichuan province as data sources and apply an object-oriented image analysis method. Firstly,the selection of parameters in multi-scale segmentation is studied,and the optimal segmentation parameters are obtained. Then,the digital eigenvalue of each image object is considered,and the improved SEaTH algorithm is used to optimize the eigenvalue optimization. Finally,the classification of low-altitude remote sensing images in the Lushan earthquake area is obtained by using the method of membership information extraction,and the classification accuracy is evaluated. The results show that the total accuracy of the classification is 87.5%,and the Kappa coefficient is 0.835. Through the study of this paper,it can provide technical support for the rapid acquisition of geospatial source data of earthquake disaster areas.