Unsupervised unmixing of hyperspectral imagery using the constrained positive matrix factorization

Abstract

This paper presents an approach for simultaneous determination of endmembers and their abundances in hyperspectral imagery unmixing using a constrained positive matrix factorization (PMF). The algorithm presented here solves the constrained PMF using Gauss-Seidel method. This algorithm alternates between the endmembers matrix updating step and the abundance estimation step until convergence is achieved. Preliminary results using a subset of a HYPERION image taken in SW Puerto Rico are presented. These results show the potential of the proposed method to solve the unsupervised unmixing problem.

Unmixing of Hyperspectral Images using Bayesian Non-negative Matrix Factorization with Volume Prior

“…In a reflectance spectroscopy system, which we use in the experiments in this paper, non-linear mixing can occur, when incident light interacts with several constituent materials. For simplicity, however, it is reasonable to assume [16] that the mixing process is predominantly linear and that non-linear effects can be neglected.…”

Bayesian nonnegative Matrix Factorization with volume prior for unmixing of hyperspectral images

“…In a reflectance acquisition system the mixing can be considered linear according to Lambert-Beer's law. Non-linear mixing are usually neglected due to minor significance and increased modeling complexity [2]. Figure 1 illustrates a hyperspectral image of a wheat kernel with a corresponding pre-processed spectrum from 950 − 1650nm.…”