基于Landsat数据的高寒湿地信息提取方法研究  

Extraction Method of Alpine Wetland Information Using Landsat Data

作　　者：陈璐 李旺平 郝君明[1,2] 周兆叶 张秀霞[1,2] 程小强 汪孝贤[1,2] CHEN Lu;LI Wangping;HAO Junming;ZHOU Zhaoye;ZHANG Xiuxia;CHENG Xiaoqiang;WANG Xiaoxian(School of Civil Engineering,Lanzhou University of Technology,Lanzhou 730050,China;Gansu Engineering Research Center for Emergency Mapping,Lanzhou University of Technology,Lanzhou 730050,China)

机构地区：[1]兰州理工大学土木工程学院,甘肃兰州730050 [2]兰州理工大学甘肃省应急测绘工程研究中心,甘肃兰州730050

出　　处：《遥感技术与应用》2024年第3期708-717,共10页Remote Sensing Technology and Application

基　　金：甘肃省自然科学基金项目(22JR5RA247、20J5RA444、20J10RA179)。

摘　　要：我国高寒湿地拥有丰富的生物多样性,是单位面积生产力最高的生态系统之一,同时也是最脆弱的生态系统,准确提取高寒湿地信息不仅有利于提高湿地动态变化监测水平,还在保护湿地生态系统多样性和湿地修复等方面具有重要意义。以青藏高原东段碌曲县为研究区,针对基于遥感技术的湿地提取存在提取精度不高,不同湿地类型的最优提取特征不明确等问题,提出了一种新的湿地提取方法,该方法首先获得了研究区的光谱、指数、地形和纹理等特征因子;其次,基于最大熵模型的Jackknife测试工具和GIS相关性分析对不同特征因子进行重要性排序和相关性分析,通过比较受试者工作特征ROC曲线下面积(AUC)值对特征因子组合进行优选;最后以最优特征因子组合为输入变量,采用最大熵耦合离散粒子群(MEDPSO)方法实现了湿地的提取。结果表明:特征因子经过最大熵模型的Jackknife测试工具和GIS相关性分析优选后,湿地提取精度显著提高,所有湿地类型的用户精度均达到90%以上,总体精度和Kappa系数分别达为87.56%和0.83。China's alpine wetlands have rich biodiversity and are one of the most productive ecosystems per unit area,as well as the most vulnerable ecosystem.Accurately extracting information from alpine wetlands is not only beneficial for improving the level of wetland dynamic change monitoring,but also of great significance in protecting wetland ecosystem diversity and wetland restoration.This article takes Luqu County in the eastern section of the Qinghai Tibet Plateau as the research area.In response to the problems of low extraction accuracy and unclear optimal extraction features for different wetland types in wetland extraction based on remote sensing technology,a new wetland extraction method is proposed.Firstly,the spectral,exponential,topographic,and texture feature factors of the research area are obtained.Secondly,the Jackknife testing tool based on the maxi⁃mum entropy model and GIS correlation analysis are used to rank and analyze the importance of different feature factors.The combination of feature factors is optimized by comparing the area under the ROC curve(AUC)val⁃ues of the working characteristics of the subjects.Finally,the optimal feature factor combination is used as the input variable,and the Maximum Entropy coupled Discrete Particle Swarm Optimization(MEDPSO)method is used to achieve wetland extraction.The results showed that after the feature factors were optimized using the Jackknife test tool of the maximum entropy model and GIS correlation analysis,the accuracy of wetland extrac⁃tion was significantly improved.The user accuracy of all wetland types reached over 90%,and the overall accu⁃racy and Kappa coefficient were 87.56%and 0.83,respectively.

关 键 词：湿地提取 特征优选 最大熵模型 离散粒子群算法 甘肃碌曲 

分 类 号：P407.8[天文地球—大气科学及气象学]