摘要
该文对近20多a来基于栅格数字高程模型(DEM)提取沟谷系统的方法与技术进行了综述。将沟谷系统提取方法归纳为三类:局部方法、全局方法、局部与全局相结合的方法,比较了各类方法的优缺点。分别综述与比较了提取沟谷系统的三个主要问题———水流路径的确定、洼地的处理以及累计汇流阈值的确定的解决算法以及存在的不足。笔者认为要使这方面的研究走向真正的实用化,必须借助其它的信息,利用人工智能和模式识别等多种手段来实现自动提取。
In this paper, the methodology of channel network extraction from grid-based digital elevation model (DEM) of the past more than 20 years is reviewed. There are mainly three methods of automated extraction of the channel networks. The first approach is the local approach which seeking to classify individual cell or connected cell regions on the basis of local surface characteristics. The second one is the global approach which attempting to accumulate more global surface information, and to build a more consistent structure of the surface features. The third approach is more complex which based on local approach and often combined with global approach. Moreover, the three most important problems of characteristic topography extraction--the assignment of flow-direction, the treatment of sink and the threshold of upstream drainage area--are reviewed.
出处
《地理与地理信息科学》
CSSCI
CSCD
北大核心
2003年第5期11-15,共5页
Geography and Geo-Information Science
基金
国家自然科学基金资助项目(40071064)
