Streetscape And POI Data-based Street Spatial Quality Evaluation

As an important part of urban public space,street space is not only the main place for residents to live,but also an important area for shaping high-quality urban space.With the return of the "people-oriented" planning concept,improving the quality of urban street space has become an important part of the current urbanisation process,which is gradually shifting towards high-quality development,and improving the quality of life of urban residents.As streetscape images can reflect the microscopic built environment of street space,and POI data can reflect the infrastructure layout of street space,and the built environment and facility layout of street space are important factors affecting the quality of street space,this thesis takes the street space of Yuzhong Peninsula as the research object,and combines streetscape and POI data to make a comprehensive evaluation of the quality of street space.The main research contents of this thesis are as follows:(1)The street space quality evaluation index system is constructed.Starting from the perception of street space users,15 evaluation indicators,such as green view rate,sky visibility factor,solar radiation intensity,vehicle traffic index,road proximity and colour richness,are selected and given specific quantification and standardisation methods to construct a street space quality evaluation index system in four perceptual dimensions,namely comfort,liveliness,safety and convenience.(2)Street function identification and street space quality evaluation index extraction.Firstly,the POI data are weighted by area and public influence,and the TF-IDF algorithm and frequency density method are used to identify street functions.Secondly,the spatial syntax method is used to analyse the road network,the semantic segmentation and coordinate conversion of street images to generate fisheye diagrams,and the spatial analysis of POI data to extract the spatial quality evaluation indicators of streets.Finally,the hierarchical analysis method is used to obtain the weights of each evaluation index.(3)An example of evaluating the spatial quality of streets in the Yuzhong Peninsula.According to the evaluation indicators and corresponding weights,the comprehensive index method was used to obtain the comprehensive results of street spatial quality,and correlation analysis was used to investigate the correlation between the evaluation indicators,population activity intensity,and street spatial quality.The results show that the spatial distribution of street quality in Yuzhong District is characterized by "large dispersion and small aggregation",with no obvious polarisation,the spatial quality of the street is distributed in the following ways:(1)Among the administrative streets.Qixinggang has the highest comprehensive quality,while Caiyuanba has the lowest,with mean values of 0.48 and 0.43 respectively.(2)Among the different functions of streets.commercial,industrial and lifestyle streets have a higher overall quality,while traffic and recreational and landscape streets have a lower overall quality.(3)In terms of road class.secondary roads have the best overall quality,followed by feeder roads and main roads.(4)In terms of the factors influencing the quality of street space.commercial vibrancy,green view rate and functional complexity are strongly positively correlated with the quality of street space,while solar radiation intensity,interface enclosure and building visibility factor are strongly negatively correlated with the quality of street space.Moran’s I = 0.187 indicates that the quality of the street space has a positive effect on the intensity of human activity,and that higher values of human activity are found in street spaces with higher activity levels.