Abstract:Current synthetic aperture radar (SAR) automatic target recognition (ATR) algorithms primarily operate under the closed-set assumption, implying that all target classes have been previously learned during the training phase. However, in open scenarios, they may encounter target classes absent from the training set, thereby necessitating an open set recognition (OSR) challenge for SAR-ATR. The crux of OSR lies in establishing distinct decision boundaries between known and unknown classes to mitigate confusion among different classes. To address this issue, we introduce a novel framework termed reinforced class separability for SAR target open set recognition (RCS-OSR), which focuses on optimizing prototype distribution and enhancing the discriminability of features. First, to capture discriminative features, a cross-modal causal features enhancement module (CMCFE) is proposed to strengthen the expression of causal regions. Subsequently, regularized intra-class compactness loss (RIC-Loss) and intra-class relationship aware consistency loss (IRC-Loss) are devised to optimize the embedding space. In conjunction with joint supervised training using cross-entropy loss, RCS-OSR can effectively reduce empirical classification risk and open space risk simultaneously. Moreover, a class-aware OSR classifier with adaptive thresholding is designed to leverage the differences between different classes. Consequently, our method can construct distinct decision boundaries between known and unknown classes to simultaneously classify known classes and identify unknown classes in open scenarios. Extensive experiments conducted on the MSTAR dataset demonstrate the effectiveness and superiority of our method in various OSR tasks.

Radar target recognition algorithm based on RCS observation sequence — set-valued identification method

Radar Hrrp Target Recognition Based On The Improved Kernel Distance Fuzzy C-Means Clustering Method

Research on Space Target Recognition Algorithm Based on Empirical Mode Decomposition

Space Target Recognition Based on RCS Feature Extraction Using Relative Position Matrix

RCS Statistical Feature Extraction for Space Target Recognition Based on Bi-LSTM

A Detection Method of Radar Signal by Wavelet Transforms

Radar Target Radar Cross-Section Measurement Based on Enhanced Imaging and Scattering Center Extraction

A New Radar Target Recognition Method Based on Wave-Coefficients

Research of Radar Automatic Target Recognition in Practical High Resolution Range Profile

Target Detection from SAR Images Based on Wavelet Transform De-Noise and Improved CFAR

Extracting Static and Dynamic Features in Joint GAF-MTF Image for Space Target Recognition

A New Radar Emitter Recognition Method Based on Variable Precision Rough Set Model

Analysis of data-driven approaches for radar target classification

A Floating Small Target Identification Method Based on Doppler Time Series Information

Study on the Application of Wavelet Transform in Target Identification of Low Resolution Radar

Radar Target Recognition by MSD Algorithms on Angular-Diversity RCS

Recognition of Targets in SAR Images Based on a WVV Feature Using a Subset of Scattering Centers

Exploring Reinforced Class Separability and Discriminative Representations for SAR Target Open Set Recognition

A Complete Automatic Target Recognition System of Low Altitude, Small RCS and Slow Speed (LSS) Targets Based on Multi-Dimensional Feature Fusion

Noise Robust Target Identification Based On The Wave-Coefficients-2dimension Case

Using Wave-Coefficients As Feature Vectors to Identify Aerospace Targets