Process flow - decompose
Decomposition (decompose) only includes principal component analys (pca) and is the last method for spectral data information enhancement (spectraInfoEnhancement). The position of the process in the chain is indicated in the schematic flow chart below.
|____SpectralData
| |____filter
| | |____singlefilter
| | |____multiFilter
| |____dataSetSplit
| | |____spectralInfoEnhancement
| | | |____scatterCorrection
| | | |____standardisation
| | | |____derivatives
| | | |____decompose
Introduction
Decomposing multi- and hyperspectral data into fewer bands or variables is oftn an efficient way to both enhance the information content and sped up processing.
he process flow decomposition only include Principal Component Analysis (PCA). The single argument required is the number of components to calculate and retain as covariates. Components will be generated from all existing input covariates and then replace these covaraites with the components.
"spectraInfoEnhancement": {
"apply": true,
"pcaPreproc": {
"apply": true,
"n_components": 8
}
}
Figure 1 illustrates decomposition of:
- original spectral reflectance,
- L2-normalised spectral reflectance, and,
- derivatives of L2-normalised spectral reflectance
