A new method for change analysis of multi-temporal hyperspectral images

In this paper, we propose a new method based on tensor factorization (TF) for hyperspectral change detection. The multilinear algebra is used to consider the whole data of multi-temporal images. The tensor-based representation and analysis has the advantage of keeping the spatial, spectral, and temporal structures in the original images. The preliminary result shows that this new method is capable of finding the foreground changes of interest in the presence of diurnal and seasonal variations.