hsdar (version 0.4.1)

bdri: Band depth ratio indices

Description

Calculate band depth ratio indices for objects of class Specfeat.

Usage

bdri(x, fnumber, index = "ndbi")

Arguments

x

Object of class Specfeat.

fnumber

Integer. Index of feature to modify.

index

Method to be applied. Currently, "bdr", "ndbi" and "bna" are available.

Value

• Object of class specfeat containing the updated version of x.

Details

Method "bdr" calculates the normalised band depth ratio as $$bdr=\frac{BD}{Dc},$$ with $BD$ is the band depth calculated by transformSpeclib and Dc is the maximum band depth called band centre. Method "ndbi" calculates the the normalised band depth index as $$ndbi= \frac{BD-Dc}{BD+Dc}.$$ Method "bna" calculates the band depth normalised to band area as $$bna=\frac{BD}{Da},$$ where $Da$ is the area of the absorption feature (see feature_properties). For further information see Mutanga and Skidmore (2004).

References

Mutanga, O. and Skidmore, A. (2004): Hyperspectral band depth analysis for a better estimation of grass biomass (Cenchrus ciliaris) measured under controlled laboratory conditions. International Journal of applied Earth Observation and Geoinformation, 5, 87-96

See Also

transformSpeclib, define.features, specfeat

Examples

data(spectral_data)

## Transform speclib
bd <- transformSpeclib(subset(spectral_data, season == "summer"),
                       method = "sh", out = "bd")

## Define features automatically
features <- define.features(bd)

## Isolate the features around 450nm, 700nm, 1200nm and 1500nm and 
## convert to specfeat.
featureSelection <- specfeat(features, c(450,700,1200,1500))

## Plot features
plot(featureSelection,1:4)

## Calculate normalized band depth index for first feature
featureSelection_bdri <- bdri(featureSelection, 1, index = "ndbi")

## Plot result
plot(featureSelection_bdri)