Linear Spectral Unmixing: Fractions Under a Model
Linear unmixing models a pixel spectrum as a weighted combination of endmember spectra.
THE ITSENSING JOURNAL
Practical satellite reviews, remote sensing methods and GIS workflows. Explore the science, understand the limits, and take your next project into iT Sensing.

Linear unmixing models a pixel spectrum as a weighted combination of endmember spectra.
An endmember is a reference spectrum used to describe a component in a spectral model.
Dimensionality reduction compresses correlated spectral channels into fewer features.
Anomaly detection finds spectra that differ from a modeled background.
Narrow-band reflectance can support models of vegetation traits such as pigments or water-related properties.
Imaging spectroscopy can detect absorption signatures associated with some atmospheric gases under suitable conditions.
A random forest combines decision trees fitted on training examples.
A support vector classifier separates labeled feature patterns using a chosen kernel and regularization.
Clustering groups observations according to feature similarity.
Spatial cross-validation separates training and evaluation according to geography or operational units.
Temporal validation evaluates a model on dates or seasons excluded from training.
A label represents a defined interpretation at a location and time.
No matching articles. Try a different search.