Abstract:Without sufficient data, the quantity of information available for supervised training is constrained, as obtaining sufficient synthetic aperture radar (SAR) training data in practice is frequently challenging. Therefore, current SAR automatic target recognition (ATR) algorithms perform poorly with limited training data availability, resulting in a critical need to increase SAR ATR performance. In this study, a new method to improve SAR ATR when training data are limited is proposed. First, an embedded feature augmenter is designed to enhance the extracted virtual features located far away from the class center. Based on the relative distribution of the features, the algorithm pulls the corresponding virtual features with different strengths toward the corresponding class center. The designed augmenter increases the amount of information available for supervised training and improves the separability of the extracted features. Second, a dynamic hierarchical-feature refiner is proposed to capture the discriminative local features of the samples. Through dynamically generated kernels, the proposed refiner integrates the discriminative local features of different dimensions into the global features, further enhancing the inner-class compactness and inter-class separability of the extracted features. The proposed method not only increases the amount of information available for supervised training but also extracts the discriminative features from the samples, resulting in superior ATR performance in problems with limited SAR training data. Experimental results on the moving and stationary target acquisition and recognition (MSTAR), OpenSARShip, and FUSAR-Ship benchmark datasets demonstrate the robustness and outstanding ATR performance of the proposed method in response to limited SAR training data.

SAR Target Recognition Via Incremental Nonnegative Matrix Factorization

SAR Target Recognition Via Incremental Nonnegative Matrix Factorization with Lp Sparse Constraint

Incremental Robust Non-negative Matrix Factorization for SAR Image Recognition

SAR Incremental Automatic Target Recognition Based on Mutual Information Maximization

SAR target recognition using nonnegative matrix factorization with L1/2 constraint

Non-Negative Matrix Factorization For Target Recognition

SAR Target Feature Extraction Based on Sparse Constraint Nonnegative Matrix Factorization

Target recognition in synthetic aperture radar images via non-negative matrix factorisation

The Feasibility Analysis of Applying NMF in SAR Target Recognition

SAR Target Recognition Based on Inception and Fully Convolutional Neural Network Combining Amplitude Domain Multiplicative Filtering Method.

SAR Target Incremental Recognition Based on Features With Strong Separability

Incremental SAR Automatic Target Recognition With Error Correction and High Plasticity

NMF and FLD Based Feature Extraction with Application to Synthetic Aperture Radar Target Recognition.

An Incremental SAR Target Recognition Framework via Memory-Augmented Weight Alignment and Enhancement Discrimination

Adaptive multiparameter spectral feature analysis for synthetic aperture radar target recognition

SAR ATR Method with Limited Training Data via an Embedded Feature Augmenter and Dynamic Hierarchical-Feature Refiner

Enhanced Synthetic Aperture Radar Automatic Target Recognition Method Based on Novel Features

Kernel Joint Sparse Representation Based SAR Automatic Target Recognition

SAR target recognition based on improved sparse LSSVM

Automatic Sar Target Recognition Based on Two-Dimensional Locality-Preserved Maximum Information Projection

Dynamic Embedding Relation Distillation Network for Incremental SAR Automatic Target Recognition