大尺度自然地理实体边界智能化提取方法OA北大核心CSTPCD

In order to solve the problem of expressing the distribution range or boundary of physical geographical objects (PGO) in maps with determinate semantics but indeterminate spatial location or distribution range,an intelligent extraction method for PGO's boundary is proposed.Firstly,the given semantic words is used to search big data of the network map.Secondly,the spatial clustering algorithm is used to determine the approximate range of PGO.Finally,considering the geometric characteristics of PGO,such as the undulations of mountains,a feature recognition algorithm is used to further determine the distribution range and boundaries of PGO.Taking into account the complexity of such objects,only the mountain (Dabie Mountains) was taken as an example to proved the effectiveness of the proposed method.