Developing urban built-up area extraction method based on land surface emissivity differences

Abstract

Accurately extracting the boundaries of urban built-up area (UBA) borders is vital for scientific urban planning and construction. Based on the land surface emissivity (LSE) and land surface temperature calculation difference between the linear spectral mixing model and Qin normalised difference vegetation index threshold improvement algorithm (ΔLST), a new method for extracting UBA is proposed. The performance of this method is assessed using Landsat images of six typical Chinese cities. The ΔLST threshold method based on the LSE difference demonstrates that the UBA boundary range is accurately extracted, and the overall average accuracy is 86.42%, which has a sufficient universality and application value. Compared with the traditional remote sensing method for extracting UBA, the method proposed in this study avoids the interference of the cloud layer, effectively solving the problem arising due to an easily extracted cloud layer, and the UBA can be accurately identified and extracted under partial cloud coverage. The proposed ΔLST threshold method that is based on the LSE difference makes up for the over-reliance on the urban heat island effect in the UBA method based on the LST inversion results.