<title>Automatic target recognition using neural networks</title>

Several researches published about artificial neural networks are connected with military problems. This research put forward ideas connected with the processing of military information to search and identify targets-automatic target recognition (ATR). A main-purpose automatic target recognition system did not exist. The research put forward here was demonstrated on military data, however it could only be considered as a proof of principle until systems were fielded and proven “under-fire”. A TR data could be in the form of non-imaging one-dimensional sensor returns, such as ultra-high range resolution radar returns for air-to-air automatic target recognition and vibration signatures from laser radar for recognition of ground targets. The ATR data could be two-dimensional images. The most common ATR images were infrared, but current systems might also deal with synthetic aperture radar images. Finally, the data could be three-dimensional, such as sequences of multiple exposures taken over time from a no stationary world.

Automatic target recognition (ATR) is a challenging task for several computer vision applications. It requires efficient, accurate, and robust methods for target detection and target identification. Deep learning has shown great success in many computer vision applications involving color RGB images. However, the performance of these networks in ATR with infrared sensor data needs further investigation. In this paper, we propose a multistage automatic target detection and recognition (ATDR) system that performs both target detection and target classification on infrared (IR) imagery using deep learning. Our system processes large IR image frames where targets take <1 % of the total number of pixels. First, we train a state-of-the-art object detector you only look once (YOLO) to localize all potential targets in the input image frame. Then, we train a convolutional neural network (CNN) to identify these detections as targets or false alarms. In this second phase, we adapt and analyze the performance of three CNN architectures: a compact and fully connected CNN, VGG16 with batch normalization, and a wide residual neural network (WRN). We also explore the use of a loss function that optimizes directly the area under the receiver operating characteristic (ROC) curve (AUC), and adapt it to our ATR application. To enhance the robustness of the proposed ATR to perturbation and variations introduced during the detection stage, we train our CNN classifiers on automatically detected targets using YOLO, in addition to ground truth bounding boxes and apply selected data augmentation techniques. To simulate real testing environments, where the spatial location of the targets within the image frame is unknown, only YOLO-detected boxes are used during validation. We evaluate our ATDR on a real benchmark dataset that includes different vehicles captured at different resolutions. Our experiments have shown that YOLO can detect most of the targets at the expense of generating a high number of false alarms. We show that the VGG-16 network with batch normalization, which is the best performing model, can correctly identify the classes of the targets, as well as classify the majority of YOLO’s false detections into an additional nontarget class. We also show that the proposed training modification to optimize an AUC-based loss function for ATR proved to be advantageous mainly in identifying difficult targets.

Multiview synthetic aperture radar (SAR) images contain much richer information for automatic target recognition (ATR) than a single-view one. It is desirable to establish a reasonable multiview ATR scheme and design effective ATR algorithm to thoroughly learn and extract that classification information, so that superior SAR ATR performance can be achieved. Hence, a general processing framework applicable for a multiview SAR ATR pattern is first given in this paper, which can provide an effective approach to ATR system design. Then, a new ATR method using a multiview deep feature learning network is designed based on the proposed multiview ATR framework. The proposed neural network is with a multiple input parallel topology and some distinct deep feature learning modules, with which significant classification features, the intra-view and inter-view features existing in the input multiview SAR images, will be learned simultaneously and thoroughly. Therefore, the proposed multiview deep feature learning network can achieve an excellent SAR ATR performance. Experimental results have shown the superiorities of the proposed multiview SAR ATR method under various operating conditions.

Automatic target recognition has been one of the hottest research in synthetic aperture radar (SAR) data processing. Noticing that popular recognition methods cannot utilize multiple features of SAR complex data, a method fused scattering center feature and deep convolutional neural network (CNN) feature is proposed in this letter. This method contains three key parts, namely, scattering center extraction and reconstruction block, CNN feature extraction block, and final feature fusion and classification block. In this process, the scattering center feature and CNN feature are fused at the level of feature maps, which retain the space information of 2-D feature maps. What is more, the proposed half end-to-end strategy realizes the automatic update of weighting parameters in feature extraction network and subnetwork, which promotes a better recognition efficiency. Experimental results on measured SAR data show that the proposed method can achieve better accuracy than other single feature-based methods and feature fusion methods.

The purpose of the paper is to extract the region of interest (ROI) from the coarse detected synthetic aperture radar (SAR) images and discriminate if the RO I contains a target or not, so as to eliminate the false alarm, and prepare for the target recognition. Th e automatic target clustering is one of the most difficult tasks in the SAR-image automatic target recognition system. The density-based spatial clustering of applications with noise (DBSCAN) relies on a density-based notion of clusters which is designed to discover clusters of arbitrary shape. DBSCAN was first used in the SAR image processing, which has many excellent features: only two insensitivity parameters (radius of neighborhood and minimum number of points) are needed; clusters of arbitrary shapes which fit in with the coarse detected SAR images can be discovered; and the cal culation time and memory can be reduced. In the multi-feature ROI discrimination scheme, we extract several target features which contain the geometry features such asthe area discriminator and Radon -transform based target profile discriminator, the distribution characteristics such as the EFF discrimina tor, and the EM scattering property such as the PPR discriminator. The synthesized judgment effectively eliminates the false alarms. Keywords: automatic target recognition, synthetic aperture radar images,

This paper smmnarizes a preliminary investigation into a space-based multi-class system-oriented High Range Resolution (HRR) Automatic Target Recognition (ATR) algorithm. Recent conflicts have highlighted the need to detect, track and recognize time critical moving ground targets in all-weather standoff (airborne and space-based) sensor platforms. Current sensor technologies based on Moving Target Indication (MTI) and/or Synthetic Aperture Radar (SAR) imaging are inadequate. While MT1 provides the ability to detect and track ground moving targets, it lacks target discrimination capability due to limited bandwidth. SAR has target discrimination capability for stationary targets, but target motion will blur the image. As a result, ATR of ground moving targets using HRR radar has recently received increased interest within the Air Force. The baseline algorithms presented in this paper are template-based mean squared error classifiers. The paper includes a brief introduction into recent HRR ATR efforts at the Air Force Research Laboratory Sensors Directorate's ATR Assessment Branch (AFRL/SNAA). Motivations and limitations for space- based sensors are discussed along with potential sensor fusion algorithms. The paper concludes by discussing the advantages of HRR ATR for feature aided

Synthetic aperture radar (SAR) are high resolution imaging radar systems. In many SAR applications classifying objects that are detected within the SAR image is important. In this paper an approach is proposed to tackle the Synthetic SAR Automatic Target Recognition (ATR) problem. The proposed scheme is based on a transfer leaning approach where three different pre-trained Convolutional Neural Networks (CNNs) are used as feature extractors in combination with a Support Vector Machine classifier (SVM). The CNNs used in this paper are AlexNet, VGG16 and GoogLeNet. The performance of these three CNNs is compared in regards to the SAR-ATR problem; where it is observed that AlexNet gives the best performance accuracy of 99.27%.

Being able to adapt all weather at all times, it has been a hot research topic that using Synthetic Aperture Radar(SAR) for remote sensing. Despite all the well-known advantages of SAR, it is hard to extract features because of its unique imaging methodology, and this challenge attracts the research interest of traditional Automatic Target Recognition(ATR) methods. With the development of deep learning technologies, convolutional neural networks(CNNs) give us another way out to detect and recognize targets, when a huge number of samples are available, but this premise is often not hold, when it comes to monitoring a specific type of ships. In this paper, we propose a method to enhance the performance of Faster R-CNN with limited samples to detect and recognize ships in SAR images.

Deep neural network-based synthetic aperture radar (SAR) automatic target recognition (ATR) systems are susceptible to attack by adversarial examples, which leads to misclassification by the SAR ATR system, resulting in theoretical model robustness problems and security problems in practice. Inspired by optical images, current SAR ATR adversarial example generation is performed in the image domain. However, the imaging principle of SAR images is based on the imaging of the echo signals interacting between the SAR and objects. Generating adversarial examples only in the image domain cannot change the physical world to achieve adversarial attacks. To solve these problems, this article proposes a framework for generating SAR adversarial examples in a 3D physical scene. First, adversarial attacks are implemented in the 2D image space, and the perturbation in the image space is converted into simulated rays that constitute SAR images through backpropagation optimization methods. The mapping between the simulated rays constituting SAR images and the 3D model is established through coordinate transformation, and point correspondence to triangular faces and intensity values to texture parameters are established. Thus, the simulated rays constituting SAR images are mapped to the 3D model, and the perturbation in the 2D image space is converted back to the 3D physical space to obtain the position and intensity of the perturbation in the 3D physical space, thereby achieving physical adversarial attacks. The experimental results show that our attack method can effectively perform SAR adversarial attacks in the physical world. In the digital world, we achieved an average fooling rate of up to 99.02% for three objects in six classification networks. In the physical world, we achieved an average fooling rate of up to 97.87% for these objects, with a certain degree of transferability across the six different network architectures. To the best of our knowledge, this is the first work to implement physical attacks in a full physical simulation condition. Our research establishes a theoretical foundation for the future concealment of SAR targets in practical settings and offers valuable insights for enhancing the attack and defense capabilities of subsequent DNNs in SAR ATR systems.

In the course of the TerraSAR-X mission, various new applications based on X-Band Synthetic Aperture Radar (SAR) data have been developed and made available as operational products or services. In this article, we elaborate on proven characteristics of TerraSAR-X that are responsible for development of operational applications. This article is written from the perspective of a commercial data and service provider and the focus is on the following applications with high commercial relevance, and varying operational maturity levels: Surface Movement Monitoring (SMM), Ground Control Point (GCP) extraction and Automatic Target Recognition (ATR). Based on these applications, the article highlights the successful transition of innovative research into sustainable and operational use within various market segments. TerraSAR-X’s high orbit accuracy, its precise radar beam tracing, the high-resolution modes, and high-quality radiometric performance have proven to be the instrument’s advanced characteristics, through, which reliable ground control points and surface movement measurements are obtained. Moreover, TerraSAR-X high-resolution data has been widely exploited for the clarity of its target signatures in the fields of target intelligence and identification. TerraSAR-X’s multi temporal interferometry applications are non-invasive and are now fully standardised autonomous tools to measure surface deformation. In particular, multi-baseline interferometric techniques, such as Persistent Scatter Interferometry (PSI) and Small Baseline Subsets (SBAS) benefit from TerraSAR-X’s highly precise orbit information and phase stability. Similarly, the instrument’s precise orbit information is responsible for sub-metre accuracy of Ground Control Points (GCPs), which are essential inputs for orthorectification of remote sensing imagery, to locate targets, and to precisely georeference a variety of datasets. While geolocation accuracy is an essential ingredient in the intelligence field, high-resolution TerraSAR-X data, particularly in Staring SpotLight mode has been widely used in surveillance, security and reconnaissance applications in real-time and also by automatic or assisted target recognition software.

Automatic target recognition (ATR) of synthetic aperture radar (SAR) images is performed on either global or local features. The global features can be extracted and classified with high efficiency. However, they lack the reasoning capability thus can hardly work well under the extended operation conditions (EOCs). The local features are often more difficult to extract and classify but they provide reasoning capability for EOC target recognition. To combine the efficiency and robustness in an ATR system, a hierarchical fusion of the global and local features is proposed for SAR ATR in this paper. As the global features, the random projection features can be efficiently extracted and effectively classified by sparse representation-based classification (SRC). The physically relevant local descriptors, i.e., attributed scattering centers (ASCs), are employed for local reasoning to handle various EOCs like noise corruption, resolution variance, and partial occlusion. A one-to-one correspondence between the test and template ASC sets is built by the Hungarian algorithm. Then, the local reasoning is performed by evaluating individual matched pairs as well as the false alarms and missing alarms. For the test image to be recognized, it is first classified by the global classifier, i.e., SRC. Once a reliable decision is made, the whole recognition process terminates. When the decision is not reliable enough, it is passed to the local classifier, where a further classification by ASC matching is carried out. Therefore, by the hierarchical fusion strategy, the efficiency of global features and the robustness of local descriptors to various EOCs can be maintained jointly in the ATR system. Extensive experiments on the moving and stationary target acquisition and recognition data set demonstrate that the proposed method achieves superior effectiveness and robustness under both SOC and typical EOCs, i.e., noise corruption, resolution variance, and partial occlusion, compared with some other SAR ATR methods.

The region of interest (ROI) contains a potential target in synthetic aperture radar (SAR) image. Picking out ROI chips before target recognizing will accelerate the whole automatic target recognition (ATR) system. This paper improves the traditional method to obtain ROI chips through constant false alarm rate (CFAR) algorithm. With the help of integral image, we reduce redundant calculation in traditional cell average CFAR (CA-CFAR) to increase its computing speed, which is denoted as the Integral Image CFAR (II-CFAR). Finally we compare the ROI extraction results of the traditional CA-CFAR and the proposed II-CFAR on the same real SAR image. The results show that the II-CFAR detector can obtain the same detection performance with the CA-CFAR detector but with a much shorter running time.

Performance of automatic target recognition (ATR) systems depends on numerous factors including the mission description, operating conditions, sensor modality, and ATR algorithm itself. Performance prediction models sensitive to these factors could be applied to ATR algorithm design, mission planning, sensor resource management, and data collection design for algorithm verification. Ideally, such a model would return measures of performance (MOPs) such as probability of detection (Pd), correct classification (Pc), and false alarm (Pfa), all as a function of the relevant predictor variables. Here we discuss the challenges of model-based and data-based approaches to performance prediction, concentrating especially on the synthetic aperture radar (SAR) modality. Our principal conclusion for model-based performance models (predictive models derived from fundamental physics- and statistics-based considerations) is that analytical progress can be made for performance of ATR system components, but that performance prediction for an entire ATR system under realistic conditions will likely require the combined use of Monte Carlo simulations, analytical development, and careful comparison to MOPs from real experiments. The latter are valuable for their high-fidelity, but have a limited range of applicability. Our principal conclusion for data-based performance models (that fit empirically derived MOPs) offer a potentially important means for extending the utility of empirical results. However, great care must be taken in their construction due to the necessarily sparse sampling of operating conditions, the high-dimensionality of the input space, and the diverse character of the predictor variables. Also the applicability of such models for extrapolation is an open question.

High-performance neural networks for synthetic aperture radar (SAR) automatic target recognition (ATR) often encounter the challenge of data scarcity. The lack of sufficient labeled SAR image datasets leads to the consideration of using simulated data to supplement the dataset. On the one hand, electromagnetic computation simulations provide high amplitude accuracy but are inefficient for large-scale datasets due to their complex computations and physical models. On the other hand, ray tracing simulations offer high geometric accuracy and computational efficiency but struggle with low amplitude correctness, hindering accurate numerical feature extraction. Furthermore, the emergence of generative adversarial networks (GANs) provides a way to generate simulated datasets, trying to balance computational efficiency with image quality. Nevertheless, the simulated SAR images generated based on random noise lack constraints, and it is also difficult to generate images that exceed the parameter conditions of the real image’s training set. Hence, it is essential to integrate physics-based simulation techniques into GANs to enhance the generalization ability of the imaging parameters. In this paper, we present the SingleScene-SAR Simulator, an efficient framework for SAR image simulation that operates under limited real SAR data. This simulator integrates rasterized shooting and bouncing rays (SBR) with cycle GAN, effectively achieving both amplitude correctness and geometric accuracy. The simulated images are appropriate for augmenting datasets in target recognition networks. Firstly, the SingleScene-SAR Simulator employs a rasterized SBR algorithm to generate radar cross section (RCS) images of target models. Secondly, a specific training pattern for cycle GAN is established to translate noisy RCS images into simulated SAR images that closely resemble real ones. Finally, these simulated images are utilized for data augmentation. Experimental results based on the constructed dataset show that with only one scene SAR image containing 30 target chips, the SingleScene-SAR Simulator can efficiently produce simulated SAR images that exhibit high similarity in both spatial and statistical distributions compared with real images. By employing simulated SAR images for data augmentation, the accuracy of target recognition networks can be consistently and significantly enhanced.

Convolutional neural network(CNN) has achieved great success in optical image recogniton, and has been applied in research of synthetic aperture radar(SAR) automatic target recognition(ATR) recently. However, in real SAR systems, SAR images are usually compressed for transmission due to the limited wireless bandwidth, and few researches have evaluated the influence of image compression on SAR ATR task. In this paper, an efficient CNN architecture based on inception module and batch normalization layer is proposed for SAR ATR tasks, and the impact of image compression on SAR ATR is evaluated based on the proposed model. The experiments are based on MSTAR dataset, and the test images are compressed by Set Partitioning in Hierarchical Trees(SPIHT) algorithm with different compression ratio. Experimental results show that the proposed CNN model achieves a state-of-the-art classification accuracy of 99.29% on original MSTAR dataset, and can still get high classification accuracy above 90% even SAR images in test set are compressed by nearly one hundred times, which reveals that moderate compression of SAR images has little influence on SAR ATR tasks, and this validates the practicability of applying the proposed model to real SAR ATR systems.