Enhancing the spatial resolution of Landsat data for mapping urban areas
Publication Date
July 19, 2017
Creator
Abstract
Detailed land cover information is crucial for mapping and managing complex urban environments across local and regional scales (Zhou and Qiu, 2015). This thesis is based on the proposition that spatial resolution is the most influential factor when mapping complex urban environments, compared to imagery’s spectral properties and type of classifier. As such, the modern “very high resolution” sensors (i.e., WorldView-2) offer a significant advantage for mapping, however using such imagery is a costly and resource-hungry approach. The coarser resolution of Landsat data (30m) is the key limitation for using these data, yet they are free and now have a temporal legacy. This doctoral research assesses the potential of using an approach that enhances the spatial resolution of Landsat data for urban land cover mapping, namely sparse representation. Focusing on the land cover mapping of the urban area of Nottingham, UK, and after establishing the superior role of spatial resolution on the accuracy of that mapping, this research demonstrates the potential of this approach. Moreover, some parameters around its use are established, in particular, the transferability of this method over space and time. It should be noted the potential of sparse representation can be even more significant by using finer spatial resolution products (i.e., Sentinel-2 and SPOT with 10m). This reaffirmed the importance of the spatial resolution for urban land cover mapping. Then it presents the sparse representation as a successful method to enhance the spatial resolution of Landsat data for urban land cover mapping.
Item Type
ethesis
Thesis Type
PhD
Supervisors
Subjects (LC)
Associated Schools / Departments
School of Geography
eprints ID
43462
UoN Repository URI
Except where otherwise noted, this item's license is described as
File(s)![Thumbnail Image]()
Name
R.Momeni Doctoral Thesis June 2017.pdf
Type
Full-text
Description
Examined
Size
4.99 MB
Format
Adobe PDF
Checksum (MD5)
4caa5e9bb8155a501f98027d9ebabda6