提出了基于纵坡面统计特征的火星地貌类型单元划分体系,详细阐述了地形特征因子及分类所需特征空间的选取建立过程。实验结果表明,以此为基础的火星表面地貌自动分类方法基本能与高程、影像信息反映事实相符,与目前公布的分类成果具有一定的可比性,达到了预期的研究目的。
This paper analyzes the basic principles of automated landforms classification based on DTA technology for Martian landforms classification. Based on the discussion and analysis, it elaborates two important issues during the study. One is the way describing the qualities of landforms using some properties; The other one is to selecting topographic attributes and building a feature space for further research. In this paper, it describes the qualities of Martian landforms based on some statistical data generated from profiles of the study area, and it brings about three topographic attributes for feature space generation. Finally, some research results are presented as a demonstration of such approaches.