热带地理
熱帶地理
열대지리
TROPICAL GEOGRAPHY
2015年
1期
35-42
,共8页
冯家莉%刘凯%朱远辉%李勇%柳林%蒙琳
馮傢莉%劉凱%硃遠輝%李勇%柳林%矇琳
풍가리%류개%주원휘%리용%류림%몽림
无人机遥感%红树林%英罗港%面向对象分类
無人機遙感%紅樹林%英囉港%麵嚮對象分類
무인궤요감%홍수림%영라항%면향대상분류
Unmanned Aerial Vehicles%mangrove forests%Yingluo Bay%object-oriented classification
低空无人机(UAV,Unmanned Aerial Vehicles)遥感系统具有数据采集灵活、低成本且可快速获取超高分辨率影像的特色,是传统航空遥感和卫星遥感的重要补充。以广东省和广西壮族自治区交界处的英罗港港湾两侧为研究区域,将无人机遥感系统用于红树林资源的遥感调查,通过无人机航拍获取高分辨率影像,并且使用拼接的影像和目视解译方法提取红树林空间分布信息,进一步选择典型研究样地,采用面向对象的最近邻分类方法对红树林树种类型进行分类研究,并对比综述了无人机遥感和常规航空航天遥感技术对红树林资源调查监测的优缺点,无人机遥感系统非常适用于红树林资源调查。通过2 h 30 min的3架次无人机航飞工作,获取了研究区域25.29 km2的无人机影像,基于无人机影像和面向对象遥感分类方法提取的红树林空间分布信息精度超过了90%。未来无人机遥感系统将可成为调查和监测红树林资源的重要技术手段,可为相关管理部门对红树林资源的保护、管理、开发等方面的工作提供基础信息和技术支持。
低空無人機(UAV,Unmanned Aerial Vehicles)遙感繫統具有數據採集靈活、低成本且可快速穫取超高分辨率影像的特色,是傳統航空遙感和衛星遙感的重要補充。以廣東省和廣西壯族自治區交界處的英囉港港灣兩側為研究區域,將無人機遙感繫統用于紅樹林資源的遙感調查,通過無人機航拍穫取高分辨率影像,併且使用拼接的影像和目視解譯方法提取紅樹林空間分佈信息,進一步選擇典型研究樣地,採用麵嚮對象的最近鄰分類方法對紅樹林樹種類型進行分類研究,併對比綜述瞭無人機遙感和常規航空航天遙感技術對紅樹林資源調查鑑測的優缺點,無人機遙感繫統非常適用于紅樹林資源調查。通過2 h 30 min的3架次無人機航飛工作,穫取瞭研究區域25.29 km2的無人機影像,基于無人機影像和麵嚮對象遙感分類方法提取的紅樹林空間分佈信息精度超過瞭90%。未來無人機遙感繫統將可成為調查和鑑測紅樹林資源的重要技術手段,可為相關管理部門對紅樹林資源的保護、管理、開髮等方麵的工作提供基礎信息和技術支持。
저공무인궤(UAV,Unmanned Aerial Vehicles)요감계통구유수거채집령활、저성본차가쾌속획취초고분변솔영상적특색,시전통항공요감화위성요감적중요보충。이광동성화엄서장족자치구교계처적영라항항만량측위연구구역,장무인궤요감계통용우홍수림자원적요감조사,통과무인궤항박획취고분변솔영상,병차사용병접적영상화목시해역방법제취홍수림공간분포신식,진일보선택전형연구양지,채용면향대상적최근린분류방법대홍수림수충류형진행분류연구,병대비종술료무인궤요감화상규항공항천요감기술대홍수림자원조사감측적우결점,무인궤요감계통비상괄용우홍수림자원조사。통과2 h 30 min적3가차무인궤항비공작,획취료연구구역25.29 km2적무인궤영상,기우무인궤영상화면향대상요감분류방법제취적홍수림공간분포신식정도초과료90%。미래무인궤요감계통장가성위조사화감측홍수림자원적중요기술수단,가위상관관리부문대홍수림자원적보호、관리、개발등방면적공작제공기출신식화기술지지。
The emerging unmanned aerial vehicle (UAV) remote sensing is an important supplement to traditional aerial and satellite remote sensing due to its flexible, fast and cost-effective capability for acquiring very high spatial resolution imagery. In this study, we explored the application of UAV to monitoring mangrove forests. The study area is located in the Yingluo Bay, on the borders between Guangdong and Guangxi Provinces. Using the UAV, we collected aerial photos of the coastal zones of Yingluo Bay, covering an area of 25.29 km2 with three flight missions totaling 2.5 hours. The high-resolution aerial images were first mosaicked and then used for interpretation and classification of mangrove forests. A typical sample plot of the study area was selected to conduct classification of mangrove species using object-oriented classification method with the nearest neighbor classifier. The classification accuracy of visual interpretation of mangrove forests extraction and that of object-oriented nearest neighbor analysis for mangrove species classification were both higher than 90%. We also compared our approach to conventional technology of aerial and satellite remote sensing for monitoring mangrove wetlands. Results suggested that UAV would be a good choice for mangrove research. It is promised that UAV would become a popular and useful tool for researchers and government agencies to contribute to mangrove reservation and management.