基于灯光遥感和兴趣点数据的城乡结合部空间识别——以银川市为例

Spatial recognition of an urban-rural fringe based on light remote sensing and interest point data:taking Yinchuan City as an example

基于灯光亮度起伏度和兴趣点数据(POI)核密度起伏度变化特征建立特征组合值模型,对银川市城乡结合部进行空间识别.结果表明,市区-城乡结合部-乡村在夜间灯光和POI的特征组合上呈现明显的空间差异性,市区有灯光亮度波动大的“高亮度、高POI”特征组合;城乡结合部有波动较小的“中灯光与低POI”特征组合;乡村呈现变化波动小的“低灯光与低POI”特征组合.城乡结合部主要分布在主城区外延23 km左右的环状地域内,面积占银川市域的51.27%.通过空间比较法将城乡结合部识别结果、土地利用现状图与高分影像进行对比验证,城乡结合部提取准确率达88.97%.

Based on the fluctuation of the light brightness and the nuclear density of interest point data(POI),a feature combination value model was established to identify the urban-rural fringe space in Yinchuan City.The research showed that the spatial differences of nighttime light fluctuation and POI distribution between urban,urban-rural fringe and rural areas.There was a combination of’high brightness and high POI’features with large brightness fluctuation in urban areas;a combination of’medium light and low POI’features with less fluctuation in rural-urban fringe zones,and a combination of’low light and low POI’with small fluctuations in rural areas.The urban-rural fringe was mainly distributed in the circular area about 23 km outside the urban core built-up area,and the area of rural-urban fringe accounted for 51.27%of the total area of Yinchuan.Through a spatial comparison the recognition results of urban-rural fringe with the land-use data and high resolution images showed the extraction accuracy to be 88.97%.