首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于无人机遥感和面向对象法的简单地物分类研究
引用本文:袁慧洁.基于无人机遥感和面向对象法的简单地物分类研究[J].测绘与空间地理信息,2020(3):113-117,123.
作者姓名:袁慧洁
作者单位:山东科技大学测绘科学与工程学院
摘    要:近年来,全国各地进行了大范围的土地利用调查,随着无人机遥感技术越来越成熟,无人机影像分析技术已深入应用到土地利用调查中,其中最多的用途是地物分类。本文选择昭通市昭阳区某乡镇区域为研究区,对采集到的无人机影像进行预处理,生成对应的正射影像;基于多种可见光植被指数,计算每3种指数合并得到影像的OIF指数,确定最佳波段组合;采用基于规则和基于样本两种面向对象分类方法,提取房屋、道路、植被等简单地物及背景。分析结果:两种方法的提取精度均达到90%以上,基于规则的面向对象分类方法精度较高,但耗时较长;基于样本的面向对象方法耗时较短,精度相对较低。两种方法相结合的全自动分类提取是下一步研究的目标。

关 键 词:无人机遥感技术  正射影像  植被指数  面向对象

Research on Simple Object Classification Based on UAV Remote Sensing and Object-oriented Method
YUAN Huijie.Research on Simple Object Classification Based on UAV Remote Sensing and Object-oriented Method[J].Geomatics & Spatial Information Technology,2020(3):113-117,123.
Authors:YUAN Huijie
Institution:(College of Geomatics,Shandong University of Science and Technology,Qingdao 266590,China)
Abstract:With the development of remote sensing technology of unmanned aerial vehicle( UAV) becoming more and more mature,using of UAV image analysis technology for features classification has become a popular and effective means. This paper selects a township area in Zhaoyang District of Zhaotong City as the research area,and preprocesses the acquired image of the UAV to generate the corresponding orthophoto;calculate a variety of visible vegetation indices,and optimum index factor( OIF) index of three indices combined image,the index determines the optimal band combination;Rule-based and sample-based object-oriented classification methods are used to classify and extract simple objects such as houses,roads,vegetation,and backgrounds. The classification results show that the extraction accuracy of both methods is above 90%. The rule-based object-oriented classification method has higher precision,but it takes a long time. The sample-based object-oriented method takes less time and has lower precision. The automatic classification and extraction of the two methods is the is the next research goal.
Keywords:UAV remote sensing technology  orthophoto  vegetation index  object oriented
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号