A New Spatio-Temporal Ica For Multi-Temporal Endmembers Extraction And Change Trajectory Analysis

Independent component analysis (ICA) has been commonly applied to extract sources from hyperspectral images. The purpose of this paper is to study the ability of ICA in extracting spatio-temporal patterns from different hyperspectral sensors, or with different acquisition conditions and dates. Spatio-temporal ICA (stICA), which maximizes the degree of independence over space and time, is a suitable method for analyzing multi-temporal hyperspectral images and extracting local features.