Exploring The Effects Of local Environment On Population Distribution: Using Imagery Segmentation Technology And Street View

The relationships between local environment and the activity and mobility of urban population is one of the classic theoretical issues in urban studies. Yet one of major limitation of evaluating the local environment is that the current measurements are not precise enough and hard to quantify. This article applies the imagery segmentation technology and street view to measure the visual quality of local environment, which can effectively supplement the deficiencies of traditional variables. Then the spatial relationships between local environment and population distribution measured by Tencent locational big data are explored using spatial statistical regression models including spatial error model and spatial lag model. The results show that the local environment have significant effects on dynamics of urban population. The quantitative relationships between local environment and population distribution in this article can provide implications and suggestions for urban planning and street design.