Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
The aim of this paper is to assess the gain of accuracy obtained by taking into account spatial information for anomaly detection in hyperspectral imaging. A mixture of conditional vector autoregressive model, MixCVAR, is introduced for background pixels. It is exploited to construct an anomaly detector (AD) based on generalized likelihood ratio test (GLRT). In the considered detection task, this detector outperforms the SEM-RX detector [1].