The Study On Method Of Extraction Mountain Lines And Classification Of Landform Based On DEM

Automatic extraction of terrain features and classification of landform based on DEM have good prospects, they have become research focuses of DTA.As the main vein of mountainous terrain, mountain lines control the overall characteristics of the terrain and posture, have important guiding significance in the scientific use of terrain for infrastructure construction, regional planning, disaster prevention and control, military command decision-making. In this paper, a new method of extracting mountain lines is proposed based on multi-scale analysis and image fusion theory. In order to achieve the target, the DEM is multi-scale decomposed by the Gaussian kernel function, and ridge lines are extracted from each scale data. Then mountain lines which reflect the macro-terrain features are extracted by fusing multi-scale ridge lines.Chinese landform is vast and complex. Study of automatically classification of Chinese landform has great practical value in scientific knowledge and rational utilization of geomorphic features. In this paper, an auto-classification method for Chinese landform is proposed. The new method firstly establish basins and plateaus description model based on the terrain factor, which reference image processing method for image segmentation and connected component analysis. Then combined with relief amplitude and altitude which are suitable for the whole Chinese continent, plains, hills, mountains, basins and plateaus are automatic divided.Finally, the above methods were verified with China DEM.