ABSTRACT
This letter proposes a robust decision fusion strategy for synthetic aperture radar (SAR) automatic target recognition (ATR). The reliabilities of individual decisions and the consistency between different decisions are quantitatively evaluated. Then, only those decisions discriminative enough for target recognition are selected for the final decision fusion. Both the efficiency and robustness of the decision fusion methods can be enhanced using the proposed strategy. Moreover, as a general strategy, the proposed strategy can be applied to different types of decision fusions, e.g., the multi-feature decision fusion, multi-classifier decision fusion and multi-view decision fusion. Experiments are conducted on the moving and stationary target acquisition recognition (MSTAR) dataset under the standard operating condition and several typical extended operating conditions (EOCs) to validate the effectiveness and robustness of the proposed strategy.