Research On Data-driven Urban Hotspot Detection And Functional Area Identification

With the continuous advancement of the urbanization process,the urban construction model mainly based on "incremental expansion" is obviously difficult to solve the urban problems we are currently facing;while the "stock tapping" model based on refined considerations,the current The urban policy puts forward specific planning guidance,which has become the new direction of urban development in the future.Hotspot areas and functional areas in cities are an important part of the city's stock space.By clarifying the distribution of hotspot areas and the layout of functional areas,it is of great significance for optimizing the spatial structure within the city.And in the traditional urban space research,the land use type data used is updated slowly,and the city can only be expressed from a macro perspective,lacking real-time and refined research on potential tapping.In this paper,Shenzhen is selected as the experimental object of this city study,which combines the current multi-source geographic big data concepts to detect and identify the types of hot spots and functional areas in Shenzhen.The research content of this paper is mainly divided into the following three parts:(1)Based on the perspective of multi-source big data,detect and identify urban hotspot areas and urban functional areas.In the data preparation stage of this experiment,Shenzhen remote sensing image data,OSM data,land use status data,bus track point data and Shenzhen POI data were collected,and the methods of qualitative analysis and quantitative identification were carried out in Shenzhen‘s Urban research.(2)Proposed an improved kernel density theory to detect and analyze hot spots in cities.Based on the analysis of multiple types of elements in the city,the POI data of Shenzhen was screened and classified according to different hotspot types;the nuclear density analysis of the data of different types of hotspot areas was completed;by selecting different contour parameters,Extract the boundary lines of the hotspot area with significant changes in the hotspot area;and further analyze the hotspot area phenomenon in the city based on the original data.(3)A quantitative recognition method for urban functional areas based on multi-attribute learning fusion is proposed.First,classify and extract the Shenzhen POI data according to the type of functional area;then,process the frequency density and category ratio of the POI data in different functional areas;after that,create a raster fishing net map of Shenzhen,and The processed urban functional area data values are linked to the raster fishing nets in Shenzhen to generate a functional area type map of Shenzhen.Finally,with the form of remote sensing images and online maps,the recognition results of the functional areas are tested and verified.