Research On The Spatial Topology Representation And Feature Extraction Method Of Topographic Visibility Domain Based On DEM

Terrain visibility analysis is an important part of geospatial information,which uses computer geometry and computer graphics technology to solve the problem of visibility between the observation point and the target point on the terrain.The spatial feature based on the viewshed is an important research content in the terrain visibility analysis,it has a wide range of applications.In the fields of urban planning,landscape analysis,and military affairs,how to construct the spatial relationship of terrain based on the viewshed is the focus of research.Humans mainly perceive everything through vision.In urban planning,designers need to consider the visual beauty of the building’s viewshed.Standing in the scenic spot position,looking in a certain direction,they can see a specific landscape,then they need to take the spatial characteristics of the viewshed into consideration.Based on this,this thesis has carried out the following research work:1.A spatial topology expression method based on the viewshed is proposed.This thesis studies the spatial segmentation of the viewshed and the representation of the topological structure.The viewshed is divided into a set of sub-views.A series of attributes that represent the characteristics of the sub-views and the topological relationship between the sub-views are given,including the description attributes of the sub-view itself and the relationship between the sub-views.The description attributes of the sub-view itself include shape,size,area,density,orientation,diameter,inner diameter,average distance,sum of center distance,and so on.The relationship between the sub-views includes the minimum distance,the maximum distance,center distance,average distance,deviation distance,etc.Analyze the sparseness of the viewshed distribution and the spatial distribution between sub-views through these attributes and relationships,and establish a qualitative and quantitative description model and method for the viewshed of the viewpoint.The spatial topology expression method proposed in this thesis is of great significance to the viewshed analysis and application research.2.A method for extracting feature elements of terrain viewshed based on DBSCAN clustering algorithm is proposed.The viewshed is divided into multiple sub-views,and different sub-view is seen in each direction.How to divide these areas is a key issue.This thesis proposes to divide the viewshed by clustering analysis method.the algorithm divides the viewshed into multiple sub-views,and calculate the spatial feature element value of each sub-view,then analyze the spatial feature distribution of the sub-views.This thesis also applies the DBSCAN algorithm to terrain feature extraction,and has achieved a great result.3.A method for extracting feature elements of terrain viewshed based on the K-Means clustering algorithm is proposed.Based on the DBSCAN viewshed division,the K-Means algorithm is used to accurately divide the terrain viewshed.K-means clustering is carried out again based on the number of clusters obtained by DBSCAN clustering algorithm,and the spatial feature elements under K-means clustering are calculated,and then analyzes the terrain view characteristics reflected by it.The fusion clustering method proposed in this thesis extracts feature elements,which makes sub-view feature extraction more advantageous than single cluster feature extraction.