首页 >  , Vol. , Issue () : -

摘要

全文摘要次数: 728 全文下载次数: 2322
引用本文:

DOI:

10.11834/jrs.20200033

收稿日期:

2020-02-10

修改日期:

2020-08-11

PDF Free   EndNote   BibTeX
基于Google Earth Engine与多源遥感数据的黑龙江流域(中国境内)沼泽湿地信息提取
宁晓刚1,2, 常文涛1,3, 王 浩1,3, 张翰超3, 朱乾德4,4
1.山东科技大学 测绘科学与工程学院;2.中国测绘科学研究院;3.中国测绘科学研究院 摄影测量与遥感研究所;4.南京水利科学研究院 水文水资源与水利科学国家重点实验室
摘要:

湿地是地球上最重要的生态系统之一,在维持全球生态环境安全等方面发挥着举足轻重的作用。由于湿地独特的水文特征,传统的湿地监测需要耗费大量的人力和财力,对于大尺度的湿地信息提取更是困难重重。随着大数据和云计算的兴起,为大尺度和长时间序列的空间数据处理提供了契机。本文基于Google Earth Engine(GEE)云平台,使用Sentinel-1合成孔径雷达(SAR)数据和Sentinel-2光学数据以及地形数据利用JM距离进行特征优选,结合随机森林算法对2018年黑龙江流域沼泽湿地进行提取。研究表明:(1)Sentinel-2红边波段和Sentinel-1雷达波段以及地形数据有利于沼泽湿地信息提取,相比植被指数和水体指数沼泽的制图精度分别提高了7.56%,5.04%,4.48%;(2)利用JM距离得到的分离度表明,红边特征>其他光学特征>地形特征>雷达特征。进行特征优选后沼泽湿地的制图和用户精度分别提高了1.45%和3.02%,最终结合随机森林算法的总体精度为91.54%,沼泽的提取精度为88.55%。本研究利用GEE云平台和多源遥感数据以及机器学习算法,能够准确、快速、高效地提取大尺度范围沼泽湿地信息,具有较大的应用潜力。

Extraction of Marsh Wetland in Heilongjiang Basin (within China) Based on Google Earth Engine and Multi-source Remote Sensing Data
Abstract:

Wetland is one of the most important ecosystems on the planet, and plays a pivotal role in maintaining global ecological environment security. Due to the unique hydrological characteristics of wetlands, traditional wetland monitoring requires a lot of manpower and financial resources, and it is even more difficult to extract large-scale wetland information. With the rise of big data and cloud computing, it provides an opportunity for large-scale and long time series spatial data processing. Based on the Google Earth Engine (GEE) cloud platform, this paper uses Sentinel-1 synthetic aperture radar (SAR) data, Sentinel-2 optical data, and terrain data to use JM distance for feature optimization, combined with random forest algorithm to extract marsh wetlands in Heilongjiang Basin in 2018. The research shows that: (1) Sentinel-2 red edge bands and Sentinel-1 radar bands and terrain data conducive to marsh wetland information extraction, compared with vegetation indexes and water indexes, the producer accuracy of marsh has increased by 7.56%, 5.04%, and 4.48%; (2) The separation obtained using JM distance shows that the red-edge features> other optical features> terrain features> radar features. After using the JM distance to select features, the producer accuracy and user accuracy of the marsh wetland increased by 1.45% and 3.02%, the overall accuracy of the combined random forest algorithm was 91.54%, and the accuracy of marsh extraction was 88%. This study uses the GEE cloud platform, multi-source remote sensing data, and machine learning algorithms to accurately, quickly, and efficiently extract large-scale marsh wetland information, which has great application potential.

本文暂时没有被引用!

欢迎关注学报微信

遥感学报交流群 分享按钮