干旱区生态环境知识资源中心(Arid): 基于国产GF-1遥感影像的水体提取方法

Arid

	基于国产GF-1遥感影像的水体提取方法
其他题名	Extraction of water information based on China-made GF-1 remote sense image
	陈文倩; 丁建丽; 李艳华; 牛增懿
来源期刊	资源科学
ISSN	1007-7588
出版年	2015
卷号	37 期号:6 页码:1166-1172
中文摘要	山区水资源信息高精度提取是西北干旱区水文水循环及水资源利用等研究领域的核心热点问题,但山区由于各种限制因素,如阴影、积雪和山体裸地等,使得高分辨率遥感监测效果不佳。本文基于高分一号遥感影像,利用单波段阈值法、NDWI与多波段法进行特克斯河流域水体信息提取,通过分析3种水体提取方法的利弊,提出单波段阈值法与构建的阴影水体指数SWI=B_1+B_2-B_4相结合的决策树水体信息提取方法,用结合实地采样得到的混淆矩阵对水体区域的整体提取结果进行精度验证对比分析。实验结果表明,单波段阈值法易受山体阴影的影响,但受积雪与裸地的影响较小;NDWI法易受山区小面积阴影影响;多波段法提取水体信息较好,但受部分阴影与薄雪影响。决策树法能有效地消除积雪和裸地等气候环境的干扰,完整地提取水体信息,总体精度为92.34%以上, Kappa系数在0.82以上。
英文摘要	In the northwest arid zone of China,among the problems of water cycle,water resource utilization and other research in the field of water scarcity,extracting water body information for montane areas is a core issue. Due to various constraints,such as shadows,snow and bare ground, high-resolution satellites have a poor ability to monitor water bodies. Here,we use GF-1 images with three methods to extract water body information for the Turks River and analyze the pros and cons of these extraction methods:proposed single-band threshold method;Shadow Water Index(SWI=B_1+B_2-B_4)phase combines to a decision tree method to extract water body information;and a confusion matrix combined with field sampling. We found that the vulnerability of single-band threshold was affected by mountain shadow,but less affected by snow and bare ground;NDWI was affected by shadow;and the multiple-band method extracted water body information better but was affected by snow. Decision tree methods can effectively eliminate interference from snow and bare ground and other climate conditions,it can completely extract water body information and the overall accuracy is more than 90% and Kappa coefficient is 0.8 or more.
中文关键词	国产高分一号 ; 水体信息提取 ; 单波段阈值法 ; 阴影水体指数 ; 决策树 ; 特克斯河
英文关键词	GF-1 extraction of water body information single-band threshold method Shadow Water Index Decision tree Turks River
语种	中文
国家	中国
收录类别	CSCD
WOS类目	GEOSCIENCES MULTIDISCIPLINARY
WOS研究方向	Geology
CSCD记录号	CSCD:5455879
来源机构	新疆大学
资源类型	期刊论文
条目标识符	http://119.78.100.177/qdio/handle/2XILL650/233856
作者单位	新疆大学资源与环境科学学院, 绿洲生态教育部重点实验室, 乌鲁木齐, 新疆 830046, 中国
推荐引用方式 GB/T 7714	陈文倩,丁建丽,李艳华,等. 基于国产GF-1遥感影像的水体提取方法[J]. 新疆大学,2015,37(6):1166-1172.
APA	陈文倩,丁建丽,李艳华,&牛增懿.(2015).基于国产GF-1遥感影像的水体提取方法.资源科学,37(6),1166-1172.
MLA	陈文倩,et al."基于国产GF-1遥感影像的水体提取方法".资源科学 37.6(2015):1166-1172.