One of the most challenging task underlying many hyperspectral imagery applications is the linear unmixing. The key to linear unmixing is to nd the set of reference substances, also called endmembers, that are representative of a given scene.
This paper presents the vertex component analysis
(VCA) a new method to unmix linear mixtures of hyperspectral sources. The algorithm is unsupervised and exploits a simple geometric fact: endmembers are vertices of a simplex. The algorithm complexity, measured in floating points operations, is O(n), where n is the sample size. The eectiveness of the proposed scheme is illustrated using simulated data.