A Multi-view SAR target recognition method based on adaptive weighted decision fusion

ABSTRACT

Synthetic aperture radar (SAR) provides high-resolution observations day and night, whose resulting images can be interpreted for different applications. For the SAR automatic target recognition (ATR) problem, this letter proposes a multi-view method based on adaptive decision fusion. The joint sparse representation (JSR) model is first employed to classify the multiple views. For the output decisions from different views, adaptive weights are determined based on Shannon entropy theory. The resulting weights are used for decision fusion to linearly combine the individual decisions from different SAR images to determine the target label. The MSTAR dataset is used for the experiments, on which both the standard operating condition (SOC) and two representative extended operating conditions (EOCs) are setup. By comparison with several state-of-the-art multi-view SAR ATR methods, the validity and robustness of the proposed method can be effectively confirmed.

KEYWORDS:

• SAR
• target recognition
• joint sparse representation
• adaptive weights
• decision fusion

Disclosure statement

No potential conflict of interest was reported by the author.