Anomaly detectors search for those pixels (anomalies) having a spectral content considerably different from the surrounding background. Unlike spectral signature based algorithms, anomaly detectors operate on the radiance received at the sensor level and do not require the step of atmospheric correction. This is a great advantage in military applications requiring real time processing. The main goal of this work is to perform a comparative analysis of different anomaly detection strategies over a real data set acquired in typical operating scenarios. An interesting aspect of this work is also given by the definition of the methodology for the comparison of the algorithm performance over a real data set.
Comparative Analysis of Hyper-spectral Detection Algorithms over Real Data
ACITO N;CORSINI, GIOVANNI;DIANI, MARCO;
2004-01-01
Abstract
Anomaly detectors search for those pixels (anomalies) having a spectral content considerably different from the surrounding background. Unlike spectral signature based algorithms, anomaly detectors operate on the radiance received at the sensor level and do not require the step of atmospheric correction. This is a great advantage in military applications requiring real time processing. The main goal of this work is to perform a comparative analysis of different anomaly detection strategies over a real data set acquired in typical operating scenarios. An interesting aspect of this work is also given by the definition of the methodology for the comparison of the algorithm performance over a real data set.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.