城市建设用地能够反映城市在地域空间上的分布形态,是衡量城市发展的重要指标.针对城市建设用地的提取,本文提出一种结合遥感影像数据和兴趣点(point of information,POI)数据的城市建设用地提取方法.首先根据珞珈一号夜间灯光数据和归一化植被指数(normalized difference vegetation index,NDVI)的亮度、纹理等信息进行多尺度面向对象检测,得到一个城市建设用地;然后使用地理探测器对POI数据进行筛选,利用筛选后的POI数据进行核密度分析提取得到另一个城市建设用地;最后通过对两种提取结果进行融合,得到精确的城市建设用地.结果表明,本方法提取的城市建设用地完整性好,与实际的城市建设用地较为吻合.该方法可为使用多源数据在市级尺度上提取城市建设用地提供参考.
Extraction of urban construction land based on remote sensing data and POI data
Urban construction land can reflect the distribution pattern of cities in geographical space and is an important indicator to measure urban development.In this paper,an urban construction land extraction method combining remote sensing image data and point of interest(POI)data is proposed.Firstly,multi-scale object-oriented detection is carried out based on information of the brightness,texture and others of Luojia 1-01 data and the normalized vegetation index(NDVI)to obtain an urban construction land.Then POI data is screened by using the geodetector,and another urban construction land is obtained by extracting the kernel density analysis of the screened POI data.Finally,by fusing the two extraction results,the accurate urban construction land is obtained.The results show that the urban construction land extracted by the method in this paper has good integrity and is relatively consistent with the actual urban construction land.This method can provide reference opinions for extracting urban construction land on a municipal scale using multi-source data.
Urban construction landLuojia 1-01Nighttime ligthObject-orientedPOI dataGeodetectorKernel density analysis
万方数据
维普
