Title:
Approaches for Characterizing Non-Linear Mixtures in Hyperspectral Imagery
Abstract:
Physics-based and kernel-based approaches are considered for generalized un-mixing of pixels in a scene. Both approaches are motivated by earlier studies that indicate non-linear mixtures in reflectance space are approximately linear in albedo space. In the first approach, reflectance is converted to single-scattering albedo and a constrained linear model is applied on the computed albedo values. The other approach uses a kernel method that seeks to capture the linear behavior of albedo in non-linear mixtures of materials. Three experiments (including laboratory and aerial collections of hyperspectral imagery) are used for validation purposes.
|