A quantitative and comparative analysis of endmember extraction algorithms from hyperspectral data

By Antonio J. Plaza, Pablo Martínez, Rosa M. Pérez, Javier Plaza. IEEE Transactions on Geoscience and Remote Sensing, 2004

652 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Atmospheric Science, Media Technology

View Paper: https://www.semanticscholar.org/paper/23d92dc4598dd7dfa78560f187fa75524584c038

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Atmospheric Science

In Media Technology

Citations in the corpus

Nonnegative matrix factorization for spectral data analysis (2006, methodology): “A thorough study and comparison of various computational methods for endmember computation, in the related context of hyperspectral unmixing can be found in the work of Plaza et al., [24].”
Joint Bayesian Endmember Extraction and Linear Unmixing for Hyperspectral Imagery (2009, background): “The interested reader is invited to consult [14] and [15] for a recent performance comparison of some standard EEAs.”
Minimum Volume Simplex Analysis: A Fast Algorithm to Unmix Hyperspectral Data (2008, background): “Given a set of mixed hyperspectral vectors, linear mixture analysis, or linear unmixing, aims at estimating the number of reference materials, also called endmembers, their spectral signatures, and their abundance fractions [1, 2, 3, 4, 5, 6].”
Constrained Nonnegative Matrix Factorization for Hyperspectral Unmixing (2009, methodology): “The other metric is the root mean square error (rmse) [41], which is used to evaluate the similarity of true (S) versus estimated (Ŝ) abundances.”
Comparing endmember selection techniques for accurate mapping of plant species and land cover using imaging spectrometer data (2012, background): “…techniques (locating spectra which represent pure endmember classes) have been extensively compared in the literature (Martinez et al., 2006; Plaza et al., 2004; Veganzones & Grana, 2008), few studies have compared different endmember selection techniques (selecting a subset of endmember…”
Sparse Unmixing of Hyperspectral Data (2011, background): “Over the last decade, several algorithms have exploited this geometrical property by estimating the “smallest” simplex set containing the observed spectral vectors [9], [10].”
Earth system science related imaging spectroscopy — an assessment (2009, methodology): “Several applications exist, however, where having the desired information calculated in (near) real-time is highly desirable (Plaza et al., 2004).”
Archetypal analysis for machine learning (2010, background): “Recently, Archetypal Analysis has found use in benchmarking and market research identifying typically extreme practices, rather than just good practices [8] as well as in the analysis of astronomy spectra [9] as an approach for the end-member extraction problem [10].”
Retrieval of subpixel Tamarix canopy cover from Landsat data along the Forgotten River using linear and nonlinear spectral mixture models (2010, methodology): “To circumvent these issues, several alternatives have been developed that involve various endmember selection strategies (Carpenter et al., 1999; Dennison & Roberts, 2003; Miao et al., 2006; Okin et al., 2001; Plaza et al., 2004; Roberts et al., 1992, 1998; Smith et al., 1994.”
Similarity-Based Unsupervised Band Selection for Hyperspectral Image Analysis (2008, background): “There are quite a few endmember extraction algorithms existing, and a review and comparative study can be found in [9].”
Recent Advances in Techniques for Hyperspectral Image Processing (2009, result): “Theparametervalues selected are in agreement with those used before in previous studies (Plaza et al., 2004).”
A variable splitting augmented Lagrangian approach to linear spectral unmixing (2009, background): “Given a set of mixed hyperspectral vectors, linear mixture analysis, or linear unmixing, aims at estimating the number of reference materials, also called endmembers, their spectral signatures, and their abundance fractions [1,2,3,4,5,6].”
Total Variation Spatial Regularization for Sparse Hyperspectral Unmixing (2012, background): “Linear spectral unmixing [2] is a standard technique for spectral mixture analysis that infers a set of pure spectral signatures, called endmembers [3], [4], and the fractions of these endmembers, called abundances [5], in each pixel of the scene.”
Exploiting manifold geometry in hyperspectral imagery (2005, background): “OR MANY years, linear mixing [10]–[15], [19], [29], [ 32 ], [43], [46] and “best band” combinations have been de facto standards in the analysis of spectral data, especially hyperspectral.”