Classification of radar targets using synthetic neural networks

Radar target classification performance of neural networks is evaluated. Time-domain and frequency-domain target features are considered. The sensitivity of the neural network algorithm to changes in network topology and training noise level is examined. The problem of classifying radar targets at unknown aspect angles is considered. The performance of the neural network algorithms is compared with that of decision-theoretic classifiers. Neural networks can be effectively used as radar target classification algorithms with an expected performance within 10 dB (worst case) of the optimum classifier     

Efficient ATR using compression

We examine various model-based automatic target recognition (MBATR) classifiers to investigate the utility of model-catalog compression realized via signal-vector quantization (VQ) and feature extraction. We specifically investigate the impact of various compression rates and common automatic target recognition (ATR) scenario variations such as noise and occlusion through simulations on high-range resolution (HRR) radar and synthetic aperture radar (SAR) data. For this data, we show that significant computational savings are possible for modest decreases in classification performance.     

GMM-based target classification for ground surveillance Doppler radar

An automatic target recognition (ATR) algorithm, based on greedy learning of Gaussian mixture model (GMM) is developed. The GMMs were obtained for a wide range of ground surveillance radar targets such as walking person(s), tracked or wheeled vehicles, animals, and clutter. Maximum-likelihood (ML) and majority-voting decision schemes were applied to these models for target classification. The corresponding classifiers were trained and tested using distinct databases of target echoes, recorded by ground surveillance radar. ML and majority-voting classifiers obtained classification rates of 88% and 96%, correspondingly. Both classifiers outperform trained human operators.     

Comparison of Two Target Classification Techniques

It has been shown that radar returns in the resonance region carry information regarding the overall dimensions and shape of targets. Two radar target classification techniques developed to utilize such returns are discussed. Both of these techniques utilize resonance region backscatter measurements of the radar cross section (RCS) and the intrinsic target backscattered phase. A target catalog used for testing the techniques was generated from measurements of the RCS of scale models of modern aircraft and naval ships using a radar range at The Ohio State University. To test the classification technique, targets had their RCS and phase taken from the data base and corrupted by errors to simulate full-scale propagation path and processing distortion. Several classification methods were then used to determine how well the corrupted measurements fit the measurement target signatures in the catalog. The first technique uses nearest neighbor (NN) algorithms on the RCS magnitude and (range corrected) phase at a number (e.g., 2, 4, or 8) of operating frequencies. The second technique uses an inverse Fourier transformation of the complex multifrequency radar returns to the time domain followed by cross correlation. Comparisons are made of the performance of the two techniques as a function of signal-to-error noise power ratio for various processing options.     

Support vector machines for SAR automatic target recognition

Algorithms that produce classifiers with large margins, such as support vector machines (SVMs), AdaBoost, etc, are receiving more and more attention in the literature. A real application of SVMs for synthetic aperture radar automatic target recognition (SAR/ATR) is presented and the result is compared with conventional classifiers. The SVMs are tested for classification both in closed and open sets (recognition). Experimental results showed that SVMs outperform conventional classifiers in target classification. Moreover, SVMs with the Gaussian kernels are able to form a local “bounded” decision region around each class that presents better rejection to confusers     

Improved target classification using optimum polarimetric SARsignatures

We present a new method for automatic target/object classification by using the optimum polarimetric radar signatures of the targets/objects of interest. The state-of-the-art in radar target recognition is based mostly either on the use of single polarimetric pairs or on the four preset pairs of orthogonal polarimetric signatures. Due to these limitations, polarimetric radar processing has been fruitful only in the area of noise suppression and target detection. The use of target separability criteria for the optimal selection of radar signal state of polarizations is addressed here. The polarization scattering matrix is used for the derivation of target signatures at arbitrary transmit and receive polarization states (arbitrary polarization inclination angles and ellipticity angles). Then, an optimization criterion that minimizes the within-class distance and maximizes the between-class metrics is used for the derivation of optimum sets of polarimetric states. The results of the application of this approach on real synthetic aperture radar (SAR) data of military vehicles are obtained. The results show that noticeable improvements in target separability and consequently target classification can be achieved by the use of the optimum over nonoptimum signatures     

应用卡尔曼滤波的机载雷达跟踪系统

本文论述将滤波理论应用于机载雷达中对单个目标进行距离、速度、方位角和高低角跟踪的多环反馈系统。首先根据目标和天线的相对运动建立控制四坐标跟踪环所需的状态矢量微分方程,然后推导相应的非线性滤波算法。最后给出计算机的模拟结果。计算机模拟的结果清晰地说明采用最佳滤波的系统性能比通常的有很大改善,并且这种瞄准轴坐标系的最佳系统对目标的随机机动是不灵敏的。 本文所讨论的方法和得出的结论可以延用到地面雷达、舰载雷达以及其他有源和无源的跟踪系统。     

Tracking with classification-aided multiframe data association

In most conventional tracking systems, only the target kinematic information from, for example, a radar or sonar or an electro-optical sensor, is used in measurement-to-track association. Target class information, which is typically used in postprocessing, can also be used to improve data association to give better tracking accuracy. The use of target class information in data association can improve discrimination by yielding purer tracks and preserving their continuity. In this paper, we present the simultaneous use of target classification information and target kinematic information for target tracking. The approach presented integrates target class information into the data association process using the 2-D (one track list and one measurement list) as well as multiframe (one track list and multiple measurement lists) assignments. The multiframe association likelihood is developed to include the classification results based on the "confusion matrix" that specifies the accuracy of the target classifier. The objective is to improve association results using class information when the kinematic likelihoods are similar for different targets, i.e., there is ambiguity in using kinematic information alone. Performance comparisons with and without the use of class information in data association are presented on a ground target tracking problem. Simulation results quantify the benefits of classification-aided data association for improved target tracking, especially in the presence of association uncertainty in the kinematic measurements. Also, the benefit of 5-D (or multiframe) association versus 2-D association is investigated for different quality classifiers. The main contribution of this paper is the development of the methodology to incorporate exactly the classification information into multidimensional (multiframe) association.     

Nonparametric Radar Extraction Using a Generalized Sign Test

A nonparametric procedure used in a constant false alarm rate (CFAR) radar extractor for detecting targets in a background of noise with unknown statistical properties is described. The detector is based on a generalization of the well-known two-sample sign test and thus requires a set of reference noise observations in addition to the set of observations being tested for signal presence. The detection performance against Gaussian noise is determined for a finite number of observations and asymptotically, for both nonfluctuating and pulse-to-pulse Rayleigh fluctuating target statistics. It is noted that the performance loss, as compared to the optimum parametric detector, depends critically on the number of reference noise observations available when the number of hits per target is not large. In the same case a much larger loss is also found for a pulse-to-pulse fluctuating target even though the asymptotic loss is the same as for a nonfluctuating target. A comparison is finally made with a detector based on the Mann-Whitney test, which usually is considered to be one of the better nonparametric procedures for the two-sample case.     

Radar target classification using doppler signatures of human locomotion models

The problem of target classification for ground surveillance Doppler radars is addressed. Two sources of knowledge are presented and incorporated within the classification algorithms: 1) statistical knowledge on radar target echo features, and 2) physical knowledge, represented via the locomotion models for different targets. The statistical knowledge is represented by distribution models whose parameters are estimated using a collected database. The physical knowledge is represented by target locomotion and radar measurements models. Various concepts to incorporate these sources of knowledge are presented. These concepts are tested using real data of radar echo records, which include three target classes: one person, two persons and vehicle. A combined approach, which implements both statistical and physical prior knowledge provides the best classification performance, and it achieves a classification rate of 99% in the three-class problem in high signal-to-noise conditions.     

Bispectral analysis of atmospheric radar signals

The bispectral analysis techniques have been applied on signals received from the back-scattered signals by the MST radar. The results of this analysis are compared with that of the conventional power spectral analysis using the Fast Fourier Transform (FFT) technique. Our results showed that there is a significant advantage in using the bispectral technique over the conventional FFT technique. This study focuses on the importance of bispectral analysis in applications to atmospheric signals.     

Modeling and performance of HF/OTH radar target classificationsystems

The effects of a class of multipath propagation channels on the performance of a over-the-horizon (OTH) radar target classification system are considered. A Rician frequency-selective fading channel model is employed to characterize the effects of the multipath propagation medium and evaluate the performance of radar target classification systems. The performance of classification algorithms that employ relative amplitude, relative phase, and absolute amplitude measurements as features is investigated. Performance estimates of the various classification algorithms for interesting sets of channel parameters are obtained by means of Monte-Carlo simulations     

Surface movement radar data processing methods for airportsurveillance

An airport surveillance function operating on surface movement radar (SMR) images is proposed and evaluated. The main contributions presented are the statistical error models of the target centroid and attributes extracted from radar images, developed and applied to the design of its main data processing blocks. Besides a multihypothesis image-to-tracks assignment method, a tracking filter using the extracted orientation and a classification scheme based on target attributes is detailed. The error models confidence and processing methods performance are demonstrated through simulation in representative scenarios     

Combination of classifiers with incomplete frames of discernment

The methods for combining multiple classifiers based on belief functions require to work with a common and complete(closed) Frame of Discernment(Fo D) on which the belief functions are defined before making their combination. This theoretical requirement is however difficult to satisfy in practice because some abnormal(or unknown) objects that do not belong to any predefined class of the Fo D can appear in real classification applications. The classifiers learnt using different attributes inform...     

Superresolution HRR ATR with high definition vector imaging

A new 1-D template-based automatic target recognition (ATR) algorithm is developed and tested on high range resolution (HRR) profiles formed from synthetic aperture radar (SAR) images of targets taken from the Moving and Stationary Target Acquisition and Recognition (MSTAR) data set. In this work, a superresolution technique known as High Definition Vector Imaging (HDVI) is applied to the HRR profiles before the profiles are passed through ATR classification. The new I-D ATR system using HDVI demonstrates significantly improved target recognition compared with previous I-D ATR systems that use conventional image processing techniques. This improvement in target recognition is quantified by improvement in probability of correct classification (PCC). More importantly, the application of HDVI to HRR profiles helps to maintain the same ATR performance with reduced radar resource requirements     

False alarm effects on estimation in multitarget trackers

An analysis of false alarm effects on tracking filter performance in multitarget track-while-scan radars, using variable correlation gates, is presented. The false alarms considered originate from noise, clutter, and crossing targets. The dimensions of the correlation gates are determined by filter prediction and measurement error variances. Track association is implanted either by means of a distance weighted average of the observations or by the nearest neighbor rule. State estimation is performed by means of a second-order discrete Kalman filter, taking into consideration random target maneuvers. Measurements are made in polar coordinates, while target dynamics are estimated in Cartesian coordinates, resulting in coupled linear filter equations. the effect of false alarms on the observation noise covariance matrix, and hence on state estimation errors, is analyzed. A computer simulation example, implementing radar target tracking with a variable correlation gate in the presence of false alarms, is discussed     

Target detection with synthetic aperture radar

Target detection with synthetic aperture radar (SAR) is considered. We derive generalized likelihood ratio (GLR) detection algorithms that may be used with SAR images that are obtained with coherent subtraction or have Gaussian distributions. We analytically compare the performance of (1) a single pixel detector, (2) a detector using complete knowledge of the target signature information and known orientation information, (3) a detector using incomplete knowledge of the target signature information and known orientation information (4) a detector using unknown target signature information and known orientation information, and (5) a detector using unknown target signature information and unknown orientation information     

Robust detection of small stochastic signals

We consider the problem of detecting a stochastic signal in white not-necessarily-Gaussian noise, using vector valued observations. The locally optimal detector is presented and its performance evaluated. The least-favorable signal spectrum and noise density (over specified classes) are found, and it is shown that the detector using these least-favorable assumptions is minimax robust. The class of spectra is that of any stochastic signal of specified power whose spectrum can be bounded from above and from below by two given positive functions. The class of densities is the ε-contamination model. We present examples of the performance achievable with the robust detector in one of these the spectral uncertainty class corresponds to the unknown Doppler shift of a radar return signal. It is demonstrated that the standard matched-filter's performance degradation with increasing Doppler shift can be avoided almost entirely through use of the robust processor     

Synthetic Aperture Radar

The general theory of side-looking synthetic aperture radar systems is developed. A simple circuit-theory model is developed; the geometry of the system determines the nature of the prefilter and the receiver (or processor) is the postfilter. The complex distributed reflectivity density appears as the input, and receiver noise is first considered as the interference which limits performance. Analysis and optimization are carried out for three performance criteria (resolution, signal-to-noise ratio, and least squares estimation of the target field). The optimum synthetic aperture length is derived in terms of the noise level and average transmitted power. Range-Doppler ambiguity limitations and optical processing are discussed briefly. The synthetic aperture concept for rotating target fields is described. It is observed that, for a physical aperture, a side-looking radar, and a rotating target field, the azimuth resolution is ?/? where ? is the change in aspect angle over which the target field is viewed, The effects of phase errors on azimuth resolution are derived in terms of the power density spectrum of the derivative of the phase errors and the performance in the absence of phase errors.