Targeted change detection for stacked multi-temporal hyperspectral image

Remote sensing change detection is very useful in earth observation and geographical study. In many applications, only a few or even one change type is of interest, that is, targeted change problem. Hyperspectral images offer abundant spectral information so as to improve change detection performance. In this paper, we propose a targeted change detection method for stacked multi-temporal hyperspectral image. Firstly, we stack multi-temporal hyperspectral image into a single file. Then, principle component analysis (PCA) is applied for dimensional reduction. Target signature is extracted from input PCs and finally, ACE is employed to detect targeted change areas. The result indicated that our proposed method is effective.