机构地区: 教育部
出 处: 《测绘科学》 2018年第5期23-32,共10页
摘 要: 水体时空分布特征对于水资源监测与应用具有重要意义,为了快速、准确、高效地获取地表水体信息,借助于遥感影像进行水体提取的方法受到国内外学者关注。该文通过梳理相关研究成果,评述水体提取的3类方法:阈值法、分类器法和自动水体提取法在模型构建、分割机理和模型适用性3方面的差异。通过进一步分析现存的各类问题,指出水体物化特征与遥感光谱特征的关系研究是提高水体提取方法普适性的一条重要途径;同时,影像提取结果评价体系的标准化和规范化,将有助于推进遥感水体提取应用范围和力度。 The spatial and temporal distribution of water is of great significance to the mo-nitoring and application of water resources.In order to obtain the surface water informati-on quickly,accurately and efficiently,the method of extracting water by remote sensing ima-ge is paid by domestic and foreign scholars.In this pater,three methods of water extracti-on are reviewed:the threshold method,the classifier method and the automatic water extr-ation method,they were summaried by three aspects:model constrution,segmentation mec-hanism and model applicability.By further analyzing the existing problems,it pointed out that the research on relationship between the water physicochemical properties and the re-mote sensing spectral characteristics is an important way to improve the universality of water extration method,and also,the standardization of the image extraction results evaluation system will be helpful to promote the application of remote sensing waterextraction.