Abstract
Spectral unmixing of urban land cover relies on representative endmember libraries. For repeated mapping of multiple cities, the use of a generic spectral library, capturing the vast spectral variability of urban areas, would constitute a more operational alternative to the tedious development of image-specific libraries prior to mapping. The size and heterogeneity of such a generic library requires an efficient pruning technique to extract site-specific spectral libraries. We propose the “Automated MUsic and spectral Separability based Endmember Selection technique” (AMUSES), which selects endmember subsets with respect to the image to be processed, while accounting for internal redundancy. Experiments on simulated hyperspectral data from Brussels (Belgium) showed that AMUSES selects more relevant endmembers compared to the conventional Iterative Endmember Selection (IES) approach. This ultimately improved mapping results (kappa increased from 0.71 to 0.83). Experiments on real HyMap data from Berlin (Germany) using a combination of libraries from different cities underlined the potential of AMUSES for handling libraries with increasing levels of generality (RMSE decreased from 0.18 to 0.15, while only using 55% of the number of spectra compared to IES). Our findings contribute to the value of generic spectral databases in the development of efficient urban mapping workflows.
Highlights
The increasing availability of hyperspectral data from airborne, but especially from upcoming satellite platforms, e.g., EnMAP [1] and HyspIRI [2], presents unprecedented potential for detailed and repeated mapping of urban areas all around the globe
Pruning algorithm clearly a higher resemblance separability) the true endmembers present algorithm clearly show a higher resemblance to the true endmembers present within the respective image blocks compared to the Iterative Endmember Selection (IES) library, the latter being the same library for all within(Figure the respective image blocks compared to the IES
In our Berlin case study, we have shown the potential of AMUSES for dealing with spectral libraries derived from various sources, thereby paving the way for the widespread adoption and integration of generic spectral libraries in spectral unmixing workflows of urban areas
Summary
The increasing availability of hyperspectral data from airborne, but especially from upcoming satellite platforms, e.g., EnMAP [1] and HyspIRI [2], presents unprecedented potential for detailed and repeated mapping of urban areas all around the globe. In order to deal with the high spatial and spectral heterogeneity typically present in these environments [3], spectral unmixing approaches are generally required for mapping urban land cover. In spectral unmixing, mixed pixels are modelled as combinations of pure material spectra (or endmembers) to retrieve subpixel land cover fractions. Examples include Multiple Endmember Spectral Mixture Analysis (MESMA [4]), the Monte Carlo. Spectral Unmixing model (AutoMCU [5]), Bayesian Spectral Mixture Analysis (BSMA [6]) and sparse unmixing [7]. These algorithms typically rely on spectral libraries, i.e., collections of pure material spectra, to capture the large spectral variability in urban areas. Either based on field measurements or through the use of supervised or unsupervised image endmember
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
