The Extraction of Urban Built-Up Areas by Integrating Night-Time Light and POI Data—A Case Study of Kunming, China

Zhang Jun,Lin Han,Yuan Xiao-Die

doi:10.1109/access.2021.3054169

Abstract

With the urban built-up area becoming one of the most prominent forms of urban expansion, accurately extracting the urban built-up area is becoming more and more important to judge the urbanization process and evaluate the urban environment. Since it is difficult to significantly improve the accuracy of single satellite data in the extraction of urban built-up areas, this study proposes a method integrating POI (Point of Interest) data and Luojia-1A data to improve the extraction accuracy of urban built-up areas. In this study, integrated Density Graph, OSTU and geometric mean are used to extract urban built-up areas respectively. The highest precision of urban built-up areas extracted before data integration is 74.3% with the highest Kappa value of 0.54; while the highest precision of urban built-up area extracted after data integration is 91.4%, with the highest Kappa value of 0.91. Therefore, it can be concluded that compared with the existing widely used night-light-based methods, this method can integrate the advantages of POI data and Luojia-1A data, that is, it can not only solve the long-term oversaturation phenomenon of night light data, but also can extract urban built-up areas in a more refined way. The method used in this study, which integrates POI data and Luojia-1A data, can not only provide a new method for urban built-up area extraction, but also can play an active guiding role in urban planning and construction.

Highlights

In recent decades, with the rapid urbanization of China and the intensification of agricultural modernization in urban built-up areas, cities have undergone unlimited expansion, and changes have occurred [1], [2]
From the perspective of the overall spatial distribution of the POI point density in the study area, the distribution of POIs in urban space clearly follows is the rule that the POI density changes from dense to sparse from the urban center to the urban edge
Compared with the latest relevant research results, the overall extraction accuracy of the global artificial impairment area (GAIA) is greater than 90% on a global scale [59], while the accuracy of the urban built-up area extracted in this study is 91.4%, closing to the overall accuracy of the relevant researchers conducted by Gong and his team, which shows that the integrated data of this study has high accuracy in extracting urban built-up areas, the extracted urban built-up areas can be applied to related research and analysis of urban spatial structure [60]

Summary

Introduction

With the rapid urbanization of China and the intensification of agricultural modernization in urban built-up areas, cities have undergone unlimited expansion, and changes have occurred [1], [2]. The extraction of urban built-up areas has become increasingly important. The urban built-up area refers to the expropriated land and the non-agricultural production construction area developed by actual construction within the municipal administrative area, which includes the concentrated part of the urban area and the urban construction land that is scattered in the suburban area but has close contact with the city. This study aims to further improve the extraction effect of urban built-up areas through data integration based on existing research

Objectives

Methods

Results

Discussion

Conclusion