摘要
青海湖位于中国西北高原地区,是中国最大的高原内陆咸水湖,同时也是世界上海拔最高的湖泊之一,其水体面积的准确提取对湖泊生态环境管理至关重要。利用遥感技术,构造了一种将改进的归一化差异水体指数(Modified Normalized Difference Water Index,MNDWI)、增强型植被指数(Enhanced Vegetation Index,EVI)和监督分类这三种方法相融合的方法,以实现对青海湖水体面积的高精度提取。比较三种单一水体提取方法和融合方法的水体提取结果、总体分类精度(Overall Accuracy,OA)和Kappa系数(KappaCoefficient,Kappa),结果表明该融合方法对比三种单一方法的提取精度有显著的提升,总体分类精度最多提高了2.9284%,Kappa系数最多提高了0.3862,表明了该融合方法能够有效地提高青海湖水体面积提取的精度。
Situated in the northwestern plateau of China,Qinghai Lake is the largest inland saltwater lake on the Tibetan Plateau and one of the highest-altitude lakes in the word.Accurate extraction of its water surface area is crucial for the management of the lake’s ecological environment.Utilizing remote sensing technology,we have developed a fusion-based method that integrates three approaches:Modified Normalized Difference Water Index(MNDWI),Enhanced Vegetation Index(EVI),and supervised classification.Thismethod aims to achieve high-precision extraction of Qinghai Lake’s water surface area.By Comparing the results of the three individual water extraction methods with the fusion-based method,including water extraction results,overall classification accuracy,and Kappa coefficient,the results demonstrate a significant enhancement in accuracy using the fusion-based method.The overall classification accuracy(OA)increased by up to 2.9284%,and the Kappa coefficient(Kappa)increased by a maximum of 0.3862,indicating the effectiveness of the fusion-based method in improving the precision of Qinghai Lake’s water surface area extraction.
作者
卢金晴
张思慧
杨光
吴云龙
LU Jinqing;ZHANG Sihui;YANG Guang;WU Yunlong(School of Information Engineering,Institute of DisasterPrevention,Sanhe 065201,China;School of Earth Science,Institute of Disaster Prevention,Sanhe065201,China;School of Geography and Information Engineering,China University ofGeosciences,Wuhan 430074,China)
出处
《城市勘测》
2024年第5期94-99,共6页
Urban Geotechnical Investigation & Surveying
基金
国家自然科学基金(42274111)。
关键词
青海湖
融合方法
水体提取
总体分类精度
Kappa系数
Qinghai Lake
fusion-based method
water extraction
overall classification accuracy
kappa coefficient
