Hyperspectral change detection with IR-MAD and initial change mask

An adaptation of the previously published iteratively reweighted multivariate alteration detection (IR-MAD) method is presented in this article in the context of hyperspectral change detection. The fact that IR-MAD transformation is invariant to linear transformations is exploited in this work. First, some change pixels are eliminated based on the principal component analysis (PCA) of the difference image and then IR-MAD method is applied to the rest of the pixels after feature reduction using the PCA. The method is demonstrated on a bitemporal hyperspectral dataset. The results show good correlation with ground truth.