地理与地理信息科学

2021, v.37(04) 45-50

[打印本页] [关闭]
本期目录(Current Issue) | 过刊浏览(Past Issue) | 高级检索(Advanced Search)

基于多源光学遥感数据的湖泊湿地分类结果一致性分析
Consistency Analysis of Classification Results for Lake Wetland Based on Multi-source Optical Remote Sensing Data

朱江涛;艾金泉;陈晓勇;汤宇豪;
ZHU Jiang-tao;AI Jin-quan;CHEN Xiao-yong;TANG Yu-hao;Key Laboratory for Digital Land and Resources of Jiangxi Province,East China University of Technology;Faculty of Geomatics,East China University of Technology;Key Laboratory of the Causes and Control of Atmospheric Pollution of Jiangxi Province,East China University of Technology;

摘要(Abstract):

多源光学遥感数据是长时序研究中的重要数据源,而其分类结果的一致性分析则是多源遥感数据应用的前提和基础。该文以鄱阳湖湿地为研究对象,采用决策树方法对Landsat-8、Sentinel-2A、GF-1、HJ-1A 4种光学影像数据进行分类,以总体精度和Kappa系数评估分类精度,并基于类型面积偏差、类型面积相关、空间叠加分析对分类结果进行一致性分析。结果表明:1)利用决策树方法对不同传感器数据进行湖泊湿地分类,总体精度均高于89%,一致性较好;2)对于具体的湿地类型面积,不同传感器分类结果基本一致,以泥沙滩涂为主,水体次之,植被最少;3)分类结果中64.30%的区域具有高度一致性,完全不一致区域占12.70%。研究成果可为多源光学遥感数据用于长时序湖泊湿地变化监测的误差分析和集成使用提供参考。
Multi-source optical remote sensing data is an important data selection in long time series research, and the consistency analysis of its classification results is the premise and basis for the use of multi-source remote sensing data.This paper took Poyang Lake wetland as the research object and classified the lake wetland by using decision tree algorithm and four kinds of optical images(Landsat-8,Sentinel-2 A,GF-1 and HJ-1 A).The overall accuracy and Kappa coefficient were used to evaluate the classification accuracy.And the consistency analysis of the classification results was carried out based on area deviation coefficient, area correlation coefficient and spatial overlay analysis.The results showed that: the overall accuracy of the classification results with various images had a good consistency higher than 89%;in the classification area analysis, the classification results of different sensors were consistent, with sediment and tidal flat as the main area, followed by water and the least was vegetation; in the spatial consistency analysis, 64.30% of the classification results in the study area were highly consistent, and the completely inconsistent areas accounted for 12.70%.The research can provide a reference for error analysis and integrated use of multi-source optical remote sensing data for long-term monitoring of lake wetland changes.

关键词(KeyWords): 湖泊湿地;影像分类;决策树;一致性
lake wetland;image classification;decision tree;consistency

Abstract:

Keywords:

基金项目(Foundation): 江西省数字国土重点实验室开放基金项目(DLLJ201917);; 江西省大气污染成因与控制重点实验室开放基金项目(AE2003);; 东华理工大学研究生科技创新基金项目(DHYC-202016)

作者(Author): 朱江涛;艾金泉;陈晓勇;汤宇豪;
ZHU Jiang-tao;AI Jin-quan;CHEN Xiao-yong;TANG Yu-hao;Key Laboratory for Digital Land and Resources of Jiangxi Province,East China University of Technology;Faculty of Geomatics,East China University of Technology;Key Laboratory of the Causes and Control of Atmospheric Pollution of Jiangxi Province,East China University of Technology;

Email:

DOI:

参考文献(References):

扩展功能
本文信息
服务与反馈
本文关键词相关文章
本文作者相关文章
中国知网
分享