Skip to Main Content
This paper introduces a novel target classification method based on the extraction of target features by using natural response related late-time electromagnetic scattered field data. In the feature extraction stage, the use of multiple signal classification (MUSIC) algorithm together with a simple but effective feature fusion approach leads to a significant reduction in the sensitivity of classification accuracy to both aspect angle variations and the signal-to-noise ratio (SNR) levels of the data. Another advantage of the proposed method is that the scattered target data is needed at only a few different target aspects in the stage of classifier design. Furthermore, real time classification within a small fraction of a second is feasible due to computational simplicity offered by this method in the final decision stage. When applied to geometrically complicated targets such as small-scale aircraft, this method provides high accuracy rates even for extremely noisy data.