Glossary

ACE : adaptive cosin/coherent estimator.

ATGP : automatic target generation process endmembers induction algorithm.

bilateral function : PWC denoising of the input signal using hard or soft kernel bilateral filtering.

CEM : constrained energy minimization.

chebyshev distance : chebyshev distance between two vectors.

convex_hull_removal function : remove the convex-hull of the signal by hull quotient

corr function : sample autocorrelation matrix of a 2D matrix, taken from the mht library.

cov function : sample covariance matrix of a 2D matrix, taken from the mht library.

FCLS : fully constrained least squares.

FFT : Fast Fourier transform

FIPPI : fast iterative pixel purity index endmembers induction algorithm.

GLRT : generalized likelihood test ratio.

HfcVd : vitual dimensionality (VD) measure.

HySime : hyperspectral signal subspace estimation.

KMeans : clusters data by trying to separate samples in n groups of equal variance.

ML : machine learning.

MNF : maximum noise fraction transform.

NFINDR : maximum (or minimum) simplex volume-based endmember extraction.

NNLS : non-negative constrained least squares.

normalize : normalizes a numpy array to be in range [0, 1].

NormXCorr : normalized cross correlation.

OSP : othogonal subspace projection.

PCA : principal component analysis.

PPI : pixel purity index algorithm for endmember finding.

SAM : spectral angle mapper.

Savitzky Golay filter : a filter that remove high frequency noise from data.

SID : spectral information divergence.

SVM : Support Vector Machines.

UCLS : unconstrained least squares abundance estimation.

Withening : decorrelate and rescale the noise in the data.