015.基于服务区划型兴趣点划定城市建成区边界方法.pdfVIP

基于服务区划型兴趣点划定城市建成区边界方法 刘凌波，吴昊，赵捷，焦洪赞 摘要：城市建成区边界是测度城市规模，划定开发边界的重要依据。在当前开放大数据环境下，呈现出多 样的城市规模测度方法：表征夜间活动强度灯光影像、表征交通路网密度与规模的道路交叉口数据、表征 人群信息联系强度的基于位置的社会网络 （LBSN）数据等。上述方法对社会经济活动表征不足，近来以反 映设施布局的兴趣点 （POI）划定建成区边界方法弥补了此缺陷，研究多以核密度等值线或设施叠加刻画边 界，边界又过于抽象失真，并未能区别对待多元丰富的POI数据。本文基于断裂点理论，提出利用服务区 划型POI数据构建服务区网络单元，基于城市分形特征以H/T断裂点 （Head/Tail Breaks）分类方法提取 小尺寸服务区形成城市边界，并将此法应用于手机基站数据。通过与灯光影像和道路交叉口划定边界比较 分析，研究总结指出此方法以人的活动定义城市，以长尾特征迭代进行科学分类，提高了城市建成区边界 精度和社会表征性。 关键词：服务区划，POI，建成区边界，H/T断裂点法，断裂点理论 ANew MethodforDelimitatingUrbanBuiltUpAreawith Service-areaBasedPOI Abstract Theurbanbuiltup areaboundary isan importantbasisfor measuringthe size ofthecity and delineating ： the development boundary. In the current open large data environment, a variety of measurement methods are presented: the DMSP/OLS data which represents the intensity of nighttime activity, the road intersection data which representsthe density and scale ofthe traffic network, and the LBSN data that characterizesthe intensity of the population’s communication, however all the methods lack representation of social economic features. Recently the POI are used to delimitate urban built up area ，utilizing kernel density and facility buffer zone. However,theoutcome abstractboundary lostresolution,notbeableto distinguish therich kindsofPOI.Therefore, this paper takes construct service area network with Service-area based POI data, extract the urban boundary by Head / Tail breaks method based on the fractal feature of cities, which method is applied on the mobile station tower as well. Based on the comparative analysis of the boundaries by the night image and the road intersection, the study concludes that this method defines the city with human activities, carries on the scientific classification with the long tail characteristic iteration, enhances the resolution of urban built up area and the representation of