摘要：Topographic feature-point groups possess the information about morphological characteristics, developing mechanism and formation process of landform. This paper takes Aster GDEM (30 × 30 m grid cells) as data source to extract peak points and saddle points in the Tibet Plateau to construct topographic feature-point groups. The quantitative depiction and spatial distribution of topographic feature-point groups in the Tibet Plateau have been studied on the basis of a series of carefully designed experiments. Based on the experiment, the classification system of landform divisions has been constructed which consists of four lowly correlated factors. After the four factors are integrated into a comprehensive layer with image integration method, automatic landform division could be realized by the method of unsupervised classification. The result shows that topographic feature-point groups analysis is an effective method in recognizing the characteristics of macro landform. The classification system of landform divisions could help distinguish plains and mountains, as well as recognize the undulation of the mountains and both overall and local uplift of the plateau. This experiment proves the great significance of the topographic feature-point method in describing landform as well as further understanding on the spatial patterns of the landform in the plateau.
关键词：the Tibet plateau;topographic feature point groups;classification;digital elevation model
