Estimate the signal to noise ratio of a hyperspectral image.

estSNR(data, p)

Arguments

data

The hyperspectral image whose signal to noise ratio needs to be estimated. Samples in rows, frequencies in columns.

p

The number of endmembers.

Value

The estimated signal to noise ratio in decibels.

Warning

Be careful, when this function is called by vca05 the data is already transposed due to lazy evaluation. If you want to get the same answer by calling this function directly, you'll need to transpose the data first!

References

Nascimento, J.M.P. and Bioucas Dias, J.M. "Vertex component analysis: a fast algorithm to unmix hyperspectral data," Geoscience and Remote Sensing, vol. 43, no. 4, pp. 898-910, April 2005, doi: 10.1109/TGRS.2005.844293