测绘学报
測繪學報
측회학보
ACTA GEODAETICA ET CARTOGRAPHICA SINICA
2014年
12期
1300-1306
,共7页
点群综合%权重%空间聚类%层次Voronoi图%树结构
點群綜閤%權重%空間聚類%層次Voronoi圖%樹結構
점군종합%권중%공간취류%층차Voronoi도%수결구
point group generalization%weight%spatial clustering%hierarchical Voronoi diagram%tree structure
通过距离权重描述点的重要程度，采用改进的k‐means算法得到点群的聚类中心，进而以聚类中心为基础，构建了层次加权Voronoi图与Voronoi层次树结构。以点群的分布范围、排列方式与密度为度量，给出了基于Voronoi层次树结构的点群综合方法，确保了点群综合前后在空间形态分布上的一致性。结合地理统计学计算，对综合方法作了进一步的量化评估与优化。经验证，本文方法是可行、有效的。
通過距離權重描述點的重要程度，採用改進的k‐means算法得到點群的聚類中心，進而以聚類中心為基礎，構建瞭層次加權Voronoi圖與Voronoi層次樹結構。以點群的分佈範圍、排列方式與密度為度量，給齣瞭基于Voronoi層次樹結構的點群綜閤方法，確保瞭點群綜閤前後在空間形態分佈上的一緻性。結閤地理統計學計算，對綜閤方法作瞭進一步的量化評估與優化。經驗證，本文方法是可行、有效的。
통과거리권중묘술점적중요정도，채용개진적k‐means산법득도점군적취류중심，진이이취류중심위기출，구건료층차가권Voronoi도여Voronoi층차수결구。이점군적분포범위、배렬방식여밀도위도량，급출료기우Voronoi층차수결구적점군종합방법，학보료점군종합전후재공간형태분포상적일치성。결합지리통계학계산，대종합방법작료진일보적양화평고여우화。경험증，본문방법시가행、유효적。
The importance of the point is described by distance weight and the clustering center point of a point group is obtained by modified k‐means algorithm .Furthermore ,the clustering center is taken as base to construct hierarchical weighted Voronoi diagram and hierarchical tree structure .Distribution scope , arrangement ,and density of the group is taken as the measurement to construct the point generalization method based on hierarchical Voronoi diagram tree structure ,thus ensuring the consistency in spatial morphology before and after .Combination with geological statistics calcul ation ,this generalization method is estimated and optimized .Finally ,the practicability and avail ability of this method is confirmed through concrete experiment .
