The increasing reliance on convolutional neural networks (CNNs) for automatic target recognition (ATR) in critical applications necessitates robust defenses against adversarial attacks, which can undermine their reliability. To address this challenge, this letter proposes a novel classification framework that enhances CNN robustness for ATR under adversarial perturbations. Although CNNs are renowned for their high recognition accuracy, their performance can be compromised by subtle adversarial perturbations designed to deceive the classifier. Our methodology is based on extracting specific features from Shapley additive explanations (SHAP) analysis within and outside the detected target area. These features are then used to train a multinomial logistic regression model using the training labels, and the trained regressor performs the classification. The key strength of our framework relies on robustness enhancement against adversarial attacks, particularly designed by the fast gradient sign method (FGSM). We validate our findings through extensive evaluations using two publicly available datasets: the multitype aircraft remote sensing images (MTARSI) dataset, which contains optical images of various aircraft types, and the moving and stationary target acquisition and recognition (MSTAR) dataset, which contains radar images.

Single-channel synthetic aperture radar (SAR) can be applied to reconstruct the 3-D structure of maritime targets with the advantages of concise hardware design and lower data rate. Nevertheless, two problems should be solved to enable it, which are: 1) 2-D radar image focusing and 2) scatterer tracking. In this article, an integral 3-D reconstruction algorithm is proposed for solving these problems. First, the radar image is refocused by using the hybrid SAR and inverse SAR (ISAR) imaging technique. Second, a novel brightest scatterer (BS) extraction method is proposed combined with a clustering algorithm, which is effective and easy to implement. Third, a fast and robust scatterer tracking algorithm (STA) is presented to obtain the complete scatterer trajectories and to effectively enable the factorization method. Moreover, the proposed STA is applicable also in the presence of numerous scatterers and of scatterer scintillation. Afterward, the factorization method is employed to reconstruct the 3-D structure of the ship target. Finally, the effectiveness of the proposed algorithm is shown with the results of simulated and real measured data.

Automatic target recognition (ATR) is one of the most demanding applications of synthetic aperture radar (SAR) in the field of radar reconnaissance and surveillance. Convolutional neural networks (CNNs) have been extensively employed for SAR-ATR and obtained remarkable accuracy. However, regarding the black-box nature and non-transparency in decision-making, CNN’s reliability is unsatisfactory. Recently, some progress has been made toward providing a visual explanation of CNN’s classification procedure. In this paper, we employ the Local Interpretable Model-agnostic Explanation (LIME) algorithm to propose an interpretability metric that can be helpful to evaluate the overall robustness of CNNs. Using the proposed framework, the user can infer what proportion of the results are based on target features, while the remainder is based on irrelevant correlations from the background clutter. The theoretical findings are validated by the public MSTAR database.

In this work, a Polarimetric Interferometric In-verse Synthetic Aperture Radar (PolInISAR) based approach is outlined for the 3-D imaging of non-cooperative targets. The role of polarimetry is proven vital here as it allows to select the optimal scattering coefficients combination through which the highest coherence can be obtained. The highest coherence reduces any chance of uncertainties in the phase estimation. Consequently, with the accurate phase, the accurate 3-D image can be reconstructed under the form of 3-D cloud. For the validation of the proposed methodology, the real Tank-72 full-polarimetric ISAR dataset is implemented. The reconstructed result is found to be better superimposing with the original Tank-72 CAD model.

This paper introduces a polarimetric three-dimensional (3-D) interferometric inverse synthetic aperture radar (ISAR) imaging process using multiple phase-centers. This approach takes effective advantage of polarimetric scattering mechanisms in 3-D target representations, which may improve target classification and identification. A Pauli decomposition scheme is considered to study the role of polarimetry in 3-D Interferometric ISAR (InISAR). The polarimetric 3-D InISAR imaging process is validated using the backhoe synthetic data released by the Air Force Research Laboratory (AFRL).

As recently demonstrated, passive radars are able to detect and track targets by exploiting illuminators of opportunity. In this paper, it will be proven that the same concept can be extended to passive inverse synthetic aperture radar (P-ISAR) imaging. A suitable type of signal processing is proposed that is able to form P-ISAR images starting from range-Doppler maps, which represent the output of passive-radar signal processing. Multiple-channel digital television broadcasting (DVB)-T signals are used to demonstrate the concept as they provide enough range resolution to form meaningful ISAR images. The problem of grating lobes, which are generated by the DVB-T signal, is also addressed and solved.

Reliable automatic target recognition (ATR) systems based on inverse synthetic aperture radar (ISAR) images require a robust feature selection. An ATR system based on polarimetric ISAR images has been recently proposed that extracts bright scatterers and uses their polarimetric signatures to define classification features. Since bright scatterers could be the results of multiple scattering, the concept of polarimetrically persistent scatterers (PPSs) has been introduced in a recent work. PPS is usually associated with single scattering mechanism and, therefore, may prove to be more robust for classification purposes. In this paper, an ATR system is defined that makes use of PPS. Furthermore, a detailed analysis is carried out to emphasize the meaning of PPSs when used for ATR.

One of the main benefits brought by the use of fully polarimetric radars is the ability to identify scattering mechanisms, which are related to the target physical properties. One of the most critical problems in synthetic range-profile reconstruction is the distortion induced by the target motion. Radial target velocity and acceleration generate second- and third-order phase terms in the received signal, which produce range shift and point-spread-function smearing. The distortions induced by the target motion produce, as a consequence, a signal-to-noise ratio loss. Recently, a method based on contrast maximization has been proposed in order to compensate target radial motions using single-polarization data. In this paper, an extension of such an algorithm is proposed that exploits fully polarimetric data in order to improve the target radial motion compensation.

Inverse synthetic aperture radar (ISAR) images are often used for classifying and recognizing targets. Moreover, the use of fully polarimetric ISAR (Pol-ISAR) images enhances classification capabilities. In this paper, the authors propose a novel automatic target recognition (ATR) technique based on the use of fully Pol-ISAR images and neural networks (NNs). In order to reduce the amount of data processed by the classifier, the brightest scattering centers are first extracted by means of the Pol-CLEAN technique, and then, their scattering matrices are decomposed using Cameron's decomposition. A classifier based on the use of multilayer perceptron NN that makes use of the features extracted from the Pol-ISAR images is then implemented. A proof-of-concept test is performed on real data acquired during a controlled experiment in an anechoic chamber.

Inverse Synthetic Aperture Radar (ISAR) images are often used for classifying and recognising targets. Moreover the use of a fully polarimentric ISAR image enhances classiication capabilities. In this paper, the authors propose a novel ATR technique based on the use of fully polarimetric ISAR images and Neural Networks. In order to reduce the amount of data processed by the classifier, the brightest scattering centres are first extracted by means of the Pol-CLEAN technique and then their scattering matrices are decomposed using Cameron's decomposition. The proposed ATR algorithm is finally tested on real data.

Scattering type classification represents a significant step toward target classification. Both the surface of Poincare's sphere and Cameron's unit disc have been used separately to represent symmetric scattering matrices and to define classification methods. In this letter, the equivalence of using the surface of Poincare's sphere and Cameron's unit disc in terms of characterization and classification of symmetric scattering types is demonstrated mathematically.

Cameron's coherent target decomposition and classification is able to represent a symmetric scatterer onto a unit disc in the complex plane, and assign it to one of the six symmetrical elemental scatterer classes. Recently, Touzi et al. proposed a variation of Cameron's method by introducing a coherent analysis. Moreover the Poincare's sphere, was used instead of the unit disc for representing symmetric scattering because it was considered a more suitable domain. The aim of this work is to demonstrate the equivalence of using Poincare's sphere domain and Cameron's unit disc, in term of characterisation and classification of symmetric scattering types.