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

基于无人机高光谱不同高度的地物快速识别研究
引用本文:黄宇,陈兴海,刘业林,黄智辉,孙梅,苏运城.基于无人机高光谱不同高度的地物快速识别研究[J].安徽农业科学,2018,46(11):170-173,209.
作者姓名:黄宇  陈兴海  刘业林  黄智辉  孙梅  苏运城
作者单位:四川双利合谱科技有限公司,四川成都,200063;北京工商大学计算机与信息工程学院,北京,100048;北京卓立汉光仪器有限公司,北京,101102
摘    要:目的]探究不同高度下地物的光谱特征变化以及不同时期、不同高度下地物的有效快速识别方法。方法]以2016年8月7日200 m高空高光谱影像和2016年9月25日100、200、300 m高空高光谱影像共4景影像为研究对象,研究不同高度下地物的高光谱特征变化以及不同时期、不同高度下地物的分类精度。结果]不同高度下,植被的光谱反射率差异显著,随着高度的升高,植被特有的特征如"绿峰"降低、"红谷"升高、"红边位置"出现"红移",在近红外范围内,光谱反射率降低;综合考虑人工参与程度、处理时间和分类精度等方面,基于ISODATA法可实现不同时期、不同高度下地物的快速识别研究。结论]该研究结果为利用无人机高光谱遥感在其他领域的应用奠定了基础。

关 键 词:无人机遥感  高光谱  不同高度  地物识别  ISODATA

Fast Identification of Ground Objects Based on Different Heights of Unmanned Aerial Veh icle(UAV)Hyper spectrum
Abstract:Ab stract Objective]The aim was to study the changes of spectral characteristics of ground objects at different heights and the method of ef-fective and rapid identification of ground objects at different altitudes and in different periods.Method]Based on the 200 m altitude hyper-spectral image on August 7,2016 and 100 m,200 m,300 m altitude hyperspectral image on September 25,2016, we studied the changes of spectral characteristics of ground objects at different heights and the classification accuracy of ground objects at different altitudes and in differ-ent periods.Result]Under different heights, vegetation spectral reflectance of significant difference, as the height increases, the vegetation characteristic features such as 'green peak' reduce,'Red Valley' elevated, 'red edge'appears 'red shift', in the near-infrared range, spectral reflectance decreased.Considering artificial participation,processing time and classification precision,and so on,we could know that based on ISODATA method could be realized rapid recognition of ground objects at different heights and different periods.Conclusion]The results provide reference for the use of unmanned aerial vehicle(UAV)hyperspectral remote sensing application in other areas.
Keywords:
本文献已被 CNKI 万方数据 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