Modified Vertex Component Analysis

Source: R/vca_lopez.R

Modified VCA algorithm that aims to reduce the algorithmic complexity of the original. Intended to be called from vca.

vca_lopez(data)

Arguments

data

Data matrix. Samples in rows, frequencies in columns. This function expects data to be already dimension-reduced (for example with vca_dr). The number of endmembers is inferred as ncol(data).

Value

A list which contains:

  • indices: indices of endmembers in extraction order.

  • projection_vectors: projection vectors (row-wise) used in each iteration, included only when debuglevel >= 1.

Both elements are returned in the same order as the endmembers are extracted.

References

Lopez, S., Horstrand, P., Callico, G.M., Lopez J.F. and Sarmiento, R., "A Low-Computational-Complexity Algorithm for Hyperspectral Endmember Extraction: Modified Vertex Component Analysis," Geoscience & Remote Sensing Letters, IEEE, vol. 9 no. 3 pp. 502-506, May 2012 doi: 10.1109/LGRS.2011.2172771