Remote sensing of burned areas via PCA, Part 2: SVD-based PCA using MODIS and Landsat data

Nikos Alexandris,Nikos Koutsias,Sandeep Gupta

doi:10.1186/s40965-017-0029-0

Nikos Alexandris, Nikos Koutsias + Show 1 more

Open Access

https://doi.org/10.1186/s40965-017-0029-0

Copy DOI

Abstract

BackgroundSingular value decomposition (SVD), as an alternative solution to principal components analysis (PCA), may enhance the spectral profile of burned areas in satellite image composites.MethodsIn this regard, we combine the pre-processing options of centering, non-centering, scaling, and non-scaling the input multi-spectral data, prior to the matrix decomposition, and treat their combinations as four different SVD-based PCA versions. Using both unitemporal and bi-temporal data sets, we test all four combinations to derive principal components. We assess the effects of the transformations based on multiresponse permutation procedures and quantify the enhanced spectral separability between burned areas and other major land cover classes via the Jeffries-Matusita metric. Lastly, we evaluate visually and numerically all principal components and select a subset of interest.ResultsThe best transformation for the subset of selected components, is the uncentered-unscaled one.ConclusionsThe results indicate that an uncentered and unscaled SVD may improve the spectral separability of burned areas in some of the higher order components.

Highlights

Singular value decomposition (SVD), as an alternative solution to principal components analysis (PCA), may enhance the spectral profile of burned areas in satellite image composites
In the article “Remote sensing of burned areas via PCA, Part 1: centering, scaling and EVD vs SVD.” [1], we present in-depth the concepts of PCA [2]; past scientific literature of PCA in remote sensing applications [3]; the link of PCA to burned area mapping [4]; the implications of centering and scaling [5]; and suggest that the uncentered-unscaled SVD-based PCA variant may further improve the spectral enhancement of burned area clusters compared to the conventional centered and EVD1-based PCA
We evaluate the outcomes of SVD considering in-depth the effects of the pre-processing transformations centering and scaling via multiresponse permutation procedures (MRPP) on samples of the land cover classes of interest; by visually inspecting the principal components; and comparing the eigen11vectors12 and eigen13values14

Summary

Introduction

Singular value decomposition (SVD), as an alternative solution to principal components analysis (PCA), may enhance the spectral profile of burned areas in satellite image composites. The pre-processing options to center and scale the image composites before the matrix decomposition, can be combined in different ways [2]. Their application influences the transformation of the spectral properties of burned area clusters. A non-centered SVD, captures in the first component greater amounts of information around the mean value of the input composite [5]. This can be advantageous in isolating burned clusters in some of the higher order components. We apply and Alexandris et al Open Geospatial Data, Software and Standards (2017) 2:21

Methods

Results

Conclusion