Application of multi seasonal LANDSAT 8 TIR BAND 10 in the extraction of built-up areas.

Abstract

Built-up area extraction from remote sensing imagery has attracted more and more attention in recent years since it has a close relationship with various applications, such as city expansion, earthquake or tsunami damage assessment, population estimation. Given the advantage of large and repetitive coverage from satellite imagery, remote sensing provides an excellent cost-effective and time-saving approach for mapping built-up lands and for understanding built-up area sprawl over time. Various approaches have been developed for extracting built-up areas via different types of remote sensing imagery in recent years but these methods are faced with a couple of challenges. Therefore, this research aimed at applying the Thermal Infrared bands of the Landsat 8 imagery in extraction of built up area as a method and also assess the spectral response of the different built up Areas in the distinctive weather seasons in Uganda. The applicability of the thermal bands of Landsat 8 Imagery gave better classification accuracy of over 80% compared to other conventional methods that is Normalised difference Built up Index and the Urban Index whose overall accuracies were below 50%. The study also revealed a similar response patterns of the different built up areas however with different reflectance’s and spectral separability in the different band of the Landsat 8 bands.