Anomaly detection and important bands selection for hyperspectral images via sparse PCA

We propose a regularised version of the classical singular value decomposition for simultaneous outliers and associated important bands selection. The contributions are twofold: First, we exploit sequential optimisation techniques in L0 formulation to obtain sparse solution of classical principal component analysis. Second, we have develop new formulation for the anomaly detection problem where the simultaneous identification of important bands can be performed. Experiments in real and simulated data are included to validate the proposed method.