Radar Recognition Research Articles

Research on the target recognition of marine surveillance radar based on ensemble learning

Research on the target recognition of marine surveillance radar based on ensemble learning

One-shot learning technique regression of reconfigurable learning network for generative modeling in interconnected imaging infrastructure

Learning visual models of object classes conventionally require hundreds or a large number of training samples. Conventional gradient-based approaches for target recognition require lot of data to be trained on and require exhaustive training with high computational expense. Hence, when a new condition or untrained data is encountered, such systems inadequately misconfigure newly learned feature sets in the trained model. This misconfigures the structure of re-learned features and is then carried out in subsequent recognition stages. Thus, a development in this scenario with low training time will allow us to fend of this disadvantage. This study presents a new automatic target recognition framework that gives the enhanced performance of target-recognition system when several imaging sensors are connected with one another. This is in contrast with traditional automatic target recognition frameworks, which utilizes one-on-one computational model over synthetic-aperture radar image-processing systems. The work comprises of a learning-based classifications strategy when dealing with sharing of learned parameters over the network to discern critical changes in target-recognition performance by utilizing a novel one-shot learning-based reconfigurable learning network for image processing platform. This upgrades the networked connected CCTV and multiview synthetic-aperture radar image object identification and recognition process.

Use of Aposteriori Information in the Implementation of Radar Recognition Systems Using Neural Network Technologies

Introduction. The current need to obtain relevant, complete and reliable information about airborne objects has led to the continuous improvement of modern radar recognition systems (MRRS) as part of control systems. The development of modern MRRS has created objective prerequisites for the use of progressive and new methods and algorithms for the processing of signals using neural networks. The use of artificial neural networks with learning ability permits expansion to include many signs of recognition by using information obtained in the process of monitoring airspace.Aim. To formulate the problem and develop proposals for the use of posterior information for airspace control in radar recognition systems using neural network technologies.Materials and methods. Based on an analysis of the structure of a unified information network, an approach was formulated to facilitate the development of MRRS based on training technologies. Using the synthesis method, examples of technical solutions were proposed, which will allow the use of modern methods and signal processing algorithms using a posteriori information generated by the control system.Results. The study identified the principles of neural network training in solving the recognition problem in the process of functioning of radio electronic equipment (REE). The technical solutions pro-posed take the functioning of the integrated radar system into account, allowing the information parameters required for training MRRS in a single information field to be obtained. It is shown that the removal of restrictions associated with the functional autonomy of REE, allows the use of posterior information in the implementation of radar recognition systems. This also allows for an increase in the number of recognition signs used in the algorithms and for the database of portraits to be replenished. Conclusion. MRRS can be developed via training by removing the restrictions associated with the autonomous functioning of RES. This allows for the situational assessment to be enhanced and management decisions to be optimised.

Adaptive convolutional network for SAR image classification

Convolutional neural network (CNN) has an outstanding performance in image classification via extracting highly effective features automatically. However, when CNN is applied to target recognition of synthetic aperture radar (SAR) images, overfitting problem exists since the lack of sufficient labelled SAR images. Hence, based on the moving and stationary target acquisition and recognition (MSTAR) benchmark dataset, the authors proposed the Adaptive-Convnet, it has a strong generalisation ability and has a good performance in different testing conditions. Besides, they also studied the transfer learning method to transfer knowledge learned from simulated SAR data to real SAR image recognition, to achieve the purpose of expanding the data set at low cost. When they substitute a part of real T72 images from MSTAR with more simulated T72 SAR images and reduce the dataset bias and enhance transferability in the last three convolution layers, the classification accuracy improved.

Simulations of the OFDM signal from distinguishing the aircraft from the chaff jamming

This study discusses the capability of the orthogonal frequency division multiplexing (OFDM) radar recognition on the aircraft in the chaff jamming, in contrast to the linear frequency modulation (LFM) signal form. The process result of the LFM signal form induces a fake target in the recognition on the aircraft in the chaff jamming, because the range and the Doppler are coupled. The process result of the OFDM signal form can separate the real target and the chaffs clearly, because the range and the Doppler are uncoupled from the signal form. The OFDM signal form is a novel and efficient way to solve the chaff jamming.

SAR Target Recognition Based on Modified Sparse Representation for Ground Safety

Recognition of synthetic aperture radar (SAR) targets is a hot topic in pattern recognition field. In the previous works, the sparse representation-based classification (SRC) is successfully used in SAR target recognition with high performance. The traditional SRC is performed on the global dictionary from the training classes. As a result, the representation capability of an individual class is not fully considered. This paper modifies the traditional SRC by performing the sparse representation over the local dictionaries formed by individual classes. In this way, the reconstruction error from one class can better reflect its representation capability as for describing the test sample. By comparing the reconstruction errors of different training classes, the target label of test sample can be classified finally. In the experiments, the MSTAR dataset is used to test the proposed method, which show the good results of the proposed method.

MIMO radar signals modulation recognition based on partial instantaneous autocorrelation spectrum

Aiming at the problem of large amount of calculation and long recognition time of traditional method of modulation identification of centralised multiple-input multiple-output (MIMO) radar signal, partial zero-delay instantaneous autocorrelation spectrum is proposed to identify the modulation of centralised MIMO radar here. In this method, the modulation identification of the centralised MIMO radar is performed by extracting the ratio of the sub-peak to the highest peak in the zero-delay transient autocorrelation spectrum of the part signal. The computational complexity of this method is low, and it identifies without additional features for coded MIMO radar signals. Here, the definition of partial instantaneous correlation spectrum is given; the feasibility of using this method to identify the modulation type of the centralised MIMO radar is analysed, and the identification process is discussed. It is showed that the proposed method can shorten the recognition time while guaranteeing the detection performance under a certain SNR in the simulation results.

Azimuthal constraint representation for synthetic aperture radar target recognition along with aspect estimation

In this paper, an azimuthal constraint representation is proposed for target recognition of synthetic aperture radar (SAR) images along with aspect angle estimation. Because of SAR azimuthal sensitivity, only those training images with similar azimuths to the test sample from the truly corresponding class are useful in the linear representation. Therefore, this study performs the linear representation in individual classes within a certain azimuth interval. The target label of the test sample is identified according to the minimum reconstruction error. Based on the best location of the azimuth interval, the aspect angle of the test sample can be simultaneously estimated using a linearly weighting algorithm. To demonstrate the validity of this method, experimental investigations are conducted on the stationary target acquisition and recognition (MSTAR) dataset under several operating conditions. In addition, the proposed method is simultaneously compared with several present SAR target recognition and aspect angle estimation methods.

基于迁移深度学习的雷达信号分选识别

基于迁移深度学习的雷达信号分选识别

HRRP-based target recognition with deep contractive neural network

One of the radar high resolution range profile (HRRP) target recognition issues is the existence of noise interference, especially for the ground target. The recognition performance of traditional shallow methods degrades as suffering from the limited capability of extracting robust and discriminative features. In this paper, a novel deep neural network called stacked denoising and contractive auto-encoder (SDCAE) is designed for millimeter wave radar HRRP recognition. To enhance the capability of learning robust structure and correlations from corrupted HRRP data, a denoising contractive auto-encoder is designed by combining the advantages of denoising auto-encoder and contractive auto-encoder. As an extension of deep auto-encoders, SDCAE inherits the advantage of enhancing the robustness of features via reducing external noise, retaining local invariance to obtain more discriminative representations of training samples. Experimental results demonstrate the superior performance of the proposed method over traditional methods, especially in noise interference condition and with few training samples.

Information indicators of radar portraits of air objects and generalized indicators of the ability of recognition systems to extract information

To improve the quality of recognition of air objects, it is proposed to use a priori information contained in reference portraits, which are formed adaptively to the conditions of observation. A decisive rule is formulated on the assignment of the observed target to the k-th group under the assumption that the signal and background are normal stationary random processes with zero mean values and the covariance matrices of portraits are known. The quality criterion of recognition is proposed, the result of which implementation is a decision with a probability not below the required Ptr. The price for the implementation of this criterion is the decision content change. For the implementation of the radar recognition system (HRD) with structural-parametric adaptation of the radar it is proposed to introduce into the system a device of the quality and control forecast, which conducts the assessment (forecast) of the amount of information and change the decisive rule of the HRD system in accordance with the received assessment. An indicator of the amount of information extracted by the recognition system from the radar portrait (RLP) is introduced, which is thought as a measure of reducing uncertainty in the decision-making process on the target group with the help of the RLR system. It is shown that the amount of extracted information depends not only on the parameters of the RLP, but also on the algorithm of its processing. The potential amount of information about the goal of the k-th information group contained in the RLP is determined, the concept of a sufficiently informative portrait with the recognition of the goals of all groups is introduced. The concepts of differential and integral contrast are formalized in the case of arbitrarily correlated RLP. The introduced concepts of differential and integral contrasts for the special case of uncorrelated RLP are extended to the General case of arbitrarily correlated RLP.

UWB radar recognition system based on HOS and SVMs

This study proposes an original ultra-wideband short-range radar (UWB-SRR) recognition system based on higher-order statistics (HOS) and support vector machines (SVMs). The main purpose of this work is to improve the road safety by implementing these techniques for detection and recognition of the uncovered road users such as pedestrians and cyclists. The combination of HOS and cell-averaging constant false alarm rate (CA-CFAR) radar detector has been proposed and investigated. The results show that a combination of HOS and CA-CFAR promises a good performance for UWB radar detector. The authors have also evaluated the performance of SVM-based target recognition system using normalised radar signature as input features. A total of 1000 signatures have been extracted for each class including pedestrian, cyclist, and car, where 50% of them have been used for the training data and the rest for the validation data. The results show that the SVM gives a good performance for the proposed system, where the recognition rates are up to 96.23, 95.25 and 97.23% for the cyclist, pedestrian and car. In the real testing performance using their scenarios, the system has successfully identified 92.77% of the right cyclist, 90.82% of the right pedestrian and 90.73% of the right car.

An Efficient and Robust Framework for SAR Target Recognition by Hierarchically Fusing Global and Local Features.

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.

Adaptive multiclass correlation filters and its applications in the time series recognition

The adaptive multiclass correlation filters (AMCF) method is proposed to exploit different kinds of features and information in a unified framework for recognition. Theoretical investigation into AMCF shows that it obtains a closed-form subsolution to constrain the optimization objective, simplifying the entire inference mechanism in the multiclass classification. The time series recognition problems, such as human action recognition and radar behavior recognition, are important yet challenging tasks. However, it is still time-consuming to acquire enough labeled training samples. AMCF is capable to exploit different kinds of features to solve the time series recognition problem. With this new correlation filters-based method, we extend the original signals and handle the insufficient training set effectively. Experiments are done on the depth image based action recognition and radar behavior recognition with a small number of training examples, including MSRAction3D, MSRGesture3D, UTD-MHAD, and radar behavior datasets. Particularly, we demonstrate that the proposed action recognition system is based on the completed local binary patterns and AMCF, and successfully achieves superior performances over the state-of-the-arts.

USING ACCUMULATION AND SIMPLE VOTING METHOD FOR RECOGNIZING CLASSES OF TARGETS BASED ON JOINT DISTRIBUTION OF PAIRS OF ECHO SIGNALS AMPLITUDES

With the development of the defense industry and aviation, there is an increasing number of different air targets. In this regard, the detection and identification of the type of target is currently a actual problem. The algorithm for recognizing object classes based on a two-dimensional probability distribution density of the correlating amplitudes is known from the recently written works. In this paper the methods of increasing the probability of correct radar target classes recognition are considered. The methods are applied with regard to recognition algorithm, which makes a decision based on two attributes (effective radar cross-section and correlation coefficient of the received echo amplitudes samples).The averaging and the simple voting methods were used for increasing the recognition reliability. In this paper is considered evaluation of efficiency of proposed methods. The results of numerical simulation showed that both methods provide the significant gain.

MEASUREMENT OF THE VARIATIONS OF THE PHASE OF A SIGNAL IS A DUALFREQUENCY RECEIVER RADIO NAVIGATION SYSTEM GLONASS, IN ORDER TO RECOGNIZE OBJECTS BASED ON THEIR RADAR PORTRAITS FOR MEDIA AND SYSTEM KKP

Proposals for constructing radio images of objects based on the synthesis of apertures are considered. The methods of radar recognition of space objects (QRs) are based on the deterministic model of the reflected signal. The deterministic model of the signal allows us to reproduce, without limitations, auxiliary signs used in the detection and recognition of targets (statistical parameters and generalized functions of signal realizations, etc.). It is necessary to obtain typical values of the parameters of the received signal for each typical structural element, and also to obtain a reliable signal reflected from the target for their comparison and classification. The necessary accuracy can only be achieved by collecting a large number of statistical information on the state of the radio path, the typical values of the signals for the reflecting elements of the structure, all this being compared and making a decision on the design of the QR. The processing of such a large amount of data can only be assigned with the help of neural networks. To increase the resolving power of measurements along the line of object movement (in azimuth), an artificial increase in the size of the antenna in the azimuth direction is used-the synthesis of the aperture. Synthesis of the aperture consists in the coherent accumulation of a series of signals reflected by the target while the object is in the antenna pattern. The range to the position of the radar relative to the moving object varies, as a result of which the phase of the reflected pulses changes. The coherent accumulation of reflected pulses during the formation of the synthesized aperture must take into account the indicated phase changes. The ionosphere is variable and the model does not always correctly describe its influence in a particular region, so monitoring the signals of the Glonass NS during radar experiments will increase the accuracy of measuring the coordinate information if the task of constructing radio images of the location objects is being posed. To obtain a high azimuth resolution, it is necessary to realize the so-called floodlight mode, when the antenna tracks the movement of the object, thereby increasing the time of its observation. Using ultra-wideband radar signals when working in floodlight mode allows to obtain a radar image with a resolution of up to 15 cm in range and 25 cm in azimuth.

Sparsity constraint nearest subspace classifier for target recognition of SAR images

This paper proposes a sparsity constraint nearest subspace classifier (SNSC) for target recognition of synthetic aperture radar (SAR) images. Unlike optical images, SAR images are highly sensitive to target azimuth. Therefore, the global dictionary collaborated by samples from different classes has high between-class correlation, which will impair the performance of sparse representation-based classification (SRC). Furthermore, even on the subspace spanned by a single class, only a small number of samples with similar azimuths to the test image are highly correlated with the test image. Thus, the linear coefficients over the subspace are actually sparse ones. Therefore, in this paper we impose sparsity constraint on nearest subspace classifier (NSC) classifier and apply it to SAR target recognition. The target label of the test sample is decided to be the class with the minimum reconstruction error. The proposed method is tested on moving and stationary target acquisition and recognition (MSTAR) dataset and compared with several state-of-the-art methods and the experimental results verify the validity and robustness of the proposed method.

Target recognition in synthetic aperture radar images via joint multifeature decision fusion

Multifeature decision fusion is an effective way to promote the performance of target recognition of synthetic aperture radar (SAR) images. This paper proposes a joint multifeature decision fusion strategy for target recognition in SAR images based on multitask compressive sensing (MtCS). The proposed method can exploit the intercorrelations among different features by enforcing the constraint on the sparsity pattern. Furthermore, the time consumption for MtCS is almost the same with that of single feature-based compressive classification, such as sparse representation-based classification. Experiments on the moving and stationary target acquisition and recognition dataset and comparison with several state-of-the-art methods demonstrate the validity of the proposed method.

Generalized synthetic aperture radar automatic target recognition by convolutional neural network with joint use of two-dimensional principal component analysis and support vector machine

Convolutional neural network (CNN), as a vital part of the deep learning research field, has shown powerful potential for automatic target recognition (ATR) of synthetic aperture radar (SAR). However, the high complexity caused by the deep structure of CNN makes it difficult to generalize. An improved form of CNN with higher generalization capability and less probability of overfitting, which further improves the efficiency and robustness of the SAR ATR system, is proposed. The convolution layers of CNN are combined with a two-dimensional principal component analysis algorithm. Correspondingly, the kernel support vector machine is utilized as the classifier layer instead of the multilayer perceptron. The verification experiments are implemented using the moving and stationary target acquisition and recognition database, and the results validate the efficiency of the proposed method.

SAR Automatic Target Recognition Based on Euclidean Distance Restricted Autoencoder

Deep learning algorithms have been introduced into target recognition of synthetic aperture radar (SAR) images for extracting deep features because of its accuracy on various recognition problems with sufficient training samples. However, applying deep structures in recognizing SAR images may suffer lack of training samples. Therefore, a deep learning method is proposed in this study based on a multilayer autoencoder (AE) combined with a supervised constraint. We bind the original AE algorithm with a restriction based on Euclidean distance to use the limited training images well. Moreover, a dropout step is added to our algorithm, which is designed to prevent overfitting caused by supervised learning. Experimental results on the MSTAR dataset demonstrate the effectiveness of the proposed method on real SAR images.