Abstract
Obtaining high-precision information on functional areas is very important in urban spatial management. It has become common to integrate social perception data and high-resolution remote sensing images to identify functional areas. Based on Point of Interest (POI) data, a scoring evaluation model is constructed to identify functional zones in built-up areas. Two weight values in the model, the POI area weight and normalized kernel density value, are used in the calculation of the impact score. Functional areas are identified by the proportion of the influence score associated with each function type. The information corresponding to 10 first-level and 29 s-level functional areas is obtained by integrating the classification results for natural attributes in high-resolution images and the social attribute functional areas in the scoring evaluation model. The overall accuracy is 87.2%, and the kappa coefficient is 0.86. The spatial structure characteristics of functional areas are analyzed based on spatial syntax, the location entropy index and so on. The main functions of each street core area and high-selectivity area are consistent with the function types, are associated with high location entropy index values and coincide with distribution centres with various functions. This study is conducive to enhancing urban spatial planning, management and decision support.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have