Low Observable Targets Research Articles

Low-observable targets detection for autonomous vehicles based on dual-modal sensor fusion with deep learning approach

Environment perception is a basic and necessary technology for autonomous vehicles to ensure safety and reliable driving. A lot of studies have focused on the ideal environment, while much less work has been done on the perception of low-observable targets, features of which may not be obvious in a complex environment. However, it is inevitable for autonomous vehicles to drive in environmental conditions such as rain, snow and night-time, during which the features of the targets are not obvious and detection models trained by images with significant features fail to detect low-observable target. This article mainly studies the efficient and intelligent recognition algorithm of low-observable targets in complex environments, focuses on the development of engineering method to dual-modal image (color–infrared images) low-observable target recognition and explores the applications of infrared imaging and color imaging for an intelligent perception system in autonomous vehicles. A dual-modal deep neural network is established to fuse the color and infrared images and detect low-observable targets in dual-modal images. A manually labeled color–infrared image dataset of low-observable targets is built. The deep learning neural network is trained to optimize internal parameters to make the system capable for both pedestrians and vehicle recognition in complex environments. The experimental results indicate that the dual-modal deep neural network has a better performance on the low-observable target detection and recognition in complex environments than traditional methods.

Cross Correlation Singularity Power Spectrum Theory and Application in Radar Target Detection Within Sea Clutters

The cross correlation power spectrum of multiple signal sequences in the singularity domain is studied in this paper. With theoretical derivation and quantitative analysis, the cross correlation singularity power spectrum (CSPS) distribution theory is proposed. Developed from correlation function (CF), spectrum CF (SCF), singularity power spectrum (SPS), and multifractal cross correlation analysis, the CSPS can be applied for the correlation analysis of multiple fractal time series. In this paper, the CSPS is rigorously derived based on SPS and SCF, and is verified with classical multifractal time series. Furthermore, a target detection method based on the proposed CSPS method is also proposed. The proposed methodology is tested on sea clutters, both with and without target, from the Ice Multiparameter Imaging X-Band radar data set. The simulation results indicate that the target detection based on CSPS performs better than conventional multifractal spectrum methods, and can achieve almost 100% detection probability of detecting low-observable targets within sea clutters.

Radar target detection method based on cross‐correlation singularity power spectrum

Recently, radar target detection based on fractal theory has received great attention. Especially in the sea clutter background, fractal modelling, fractal dimension, and multifractal spectrum are widely used in radar signal processing. Here, a novel target detection method based on cross-correlation singularity power spectrum (CSPS) is proposed. Firstly, the CSPS theory and algorithm is derived based on SPS; secondly, based on the SPS and CSPS analysis of radar sea clutter, radar target detection parameters are designed, and the detection flow and algorithm based on the CSPS method is presented in detail. The proposed methodology is tested on sea clutter, both with and without targets, from the Ice Multi-parameter Imaging X -Band radar data set. The simulation results indicate that the proposed method performs better than the conventional multifractal-based methods and traditional CFAR methods, and the detection probability of detecting low-observable targets within sea clutters is close to 100%.

Low‐observable target detection using two‐stage RFRFT

Long-time coherent integration is an effective method to implement low-observable target detection in a complex scenario. Radon-fractional Fourier transform (RFRFT) can compensate the range and Doppler migration simultaneously and obviously improve the signal-to-clutter-noise ratio (SCNR) via long-time coherent integration. However, RFRFT uses a combination of moving parameters, i.e. initial range, velocity and acceleration, to search targets, increasing the computational burden. In this study, a novel two-stage RFRFT is proposed to detect low-observable targets. In the first stage, by calculating the spectrum similarity between adjacent bins, regions of interest (ROIs) can be detected in the range–Doppler domain. The processing in the first stage reduces the global search to a limited number of ROIs. In the second stage, based on the extraction of moving parameters, a refined search is performed to realise target detection and false alarm discrimination. To verify the effectiveness of the proposed two-stage method, both a simulation scenario and real radar dataset are applied. The theoretical analysis and experiment results show that the novel method is effective for the detection of dim targets with far less computation than standard RFRFT.

Algorithms of Processing of Space-Time Signals in a Multibeam Radar during Detection and Evaluation of Parameters of Low-Observable Aerial Targets

The optimal and quasi-optimal algorithms of processing of space-time signals in a multibeam radar system intended for detection and measurement of parameters of low-observable air targets of unmanned aerial vehicles (UAVs) are considered. Algorithms for detection of targets against the background of stationary and moving surfaces are obtained. The possibility to detect and measure the parameters of targets situated at any point of specified coverage area during a short time (several fractions of a second) is shown.

Detection Performance of a Forward Scatter Radar Using a Crystal Video Detector

The forward scatter radar (FSR) configuration is especially appealing for the detection of low-observable targets since it provides well-known properties for the radar cross section, which includes enhancements with respect to the monostatic and/or moderate bistatic configurations, as well as robustness to the target material and detailed geometrical characteristics. Due to the separation of a transmitter and a receiver, it is customary to use a detection scheme based on a square-law envelope detector followed by an appropriate matched filter, which we address as a crystal video detector (CVD) based on traditional terminology. This paper provides an accurate analytical expression for the detection performance of the FSR target detection using the CVD in terms of both the probability of false alarm $(P_{{\rm{fa}}})$ and the probability of detection $(P_{{\rm{d}}})$ in order to support performance prediction and FSR system design. The derived expressions are validated by comparison to Monte Carlo simulations under two different geometrical scenarios. Moreover, the comparison of the CVD performance to an ideal optimum detector (upper bound for the FSR achievable detection performance) shows the geometrical conditions, where the CVD suffers significant detection losses so that alternative detectors should be investigated. Finally, to remove the need to operate the CVD with a fixed detection threshold, two fully adaptive detectors are derived, based on the structure of the CVD scheme, which are shown to provide a constant false alarm rate (CFAR). The performance of these CFAR detectors in terms of $P_{{\rm{fa}}}$ and $P_{{\rm{d}}}$ are provided in closed form and validated through Monte Carlo simulations, showing quite small losses with respect to the fixed-threshold CVD. The application to real data acquired by a passive FSR based on frequency-modulated signals is used to show the practical effectiveness and consistency.

Multipath Maximum Likelihood Probabilistic Multihypothesis Tracker for Low Observable Targets

In many practical scenarios with multipath propagation, one target may generate multiple detections in one scan. Proper use of multipath-induced measurements can improve the detection of very low observable (VLO) targets. In this paper, a true multitarget tracker, the joint multipath maximum likelihood probabilistic multihypothesis tracker (JMP-ML-PMHT) is proposed to address this problem. The standard ML-PMHT is extended to incorporate multipath detections and jointly track multiple VLO targets. The Cramer–Rao lower bound with multipath detections is derived. Simulation results with an over-the-horizon-radar scenario show that the JMP-ML-PMHT can detect and track multiple VLO targets by effectively utilizing the information in multipath measurements.

Multi‐frame fractional Fourier transform technique for moving target detection with space‐based passive radar

This study puts forward a moving target detection technique for space-based passive radar. The main drawback of satellite transmitters for radar applications is the very low power density on the ground level, which requires long integration times to adequately increase the signal to background ratio. To this aim, in this study a new technique is proposed relying on a hybrid coherent/non-coherent integration of the received signal in the fractional Fourier transform (FrFT) domain. After the compensation of the range walk by Keystone reformatting, a coherent integration is performed over consecutive short time intervals (frames) such that the coherence of the scattering mechanism within each frame can be assumed. Then, a proper combination of the multiple frames is performed based on the kernel of the FrFT, allowing the reliable detection of low observable targets. The effectiveness of the proposed technique has been proved by simulated analysis considering very low EIRP satellites as opportunity transmitters.

Low-observable target detection in sea clutter based on the adaptive 3D-IFS algorithm

This paper mainly studies the application of the fractal self-affine theory for weak target detection in sea clutter. In this paper, the adaptive 3D-IFS (3D-IFS: three-dimensional iterated function system) algorithm is presented and a novel weak target detection model is proposed based on the algorithm. To accurately extract the weak target from the complicated background of sea clutter, we use the radar echo model of the LFM radar and the target detection model to calculate the prediction error of radar echoes. Furthermore, based on extensive analysis and simulation, we identify the scale factor, polarization model and target state as three key factors that affect the detection performance of our proposed model. Using the real data from IPIX radar and C-band radar for simulations, we can see that the model has a significant performance improvement for weak target detection compared to the traditional 3D-IFS algorithm. The detection probability of the model reaches 70% at the SCR near −10dB, and signal processing time of the model is approximately 0.3s at the SCR near −6dB, thus it meets the requirement for low-observable (SCR>−8dB) and radial velocity about 600m/s high speed weak target detection in sea clutter.

Fourier‐Bessel transform and time–frequency‐based approach for detecting manoeuvring air target in sea‐clutter

In many applications, it may be desired to decompose a non-stationary signal into its individual components. If spectral components of the non-stationary signal do not overlap in the frequency domain then Fourier transform can be used to decompose the non-stationary signal. Fourier transform fails to decompose the non-stationary signal if its spectral components overlap in the frequency domain. In this study, the authors propose Fourier-Bessel transform and the time–frequency analysis in conjunction with the fractional Fourier transform (FB-TF) method for the separation of multi-component non-stationary signal whose components overlap in both time and/or frequency domains. The efficiency of the proposed method is compared with one of the traditional decomposition methods like EMD. The proposed approach is applied to both simulated and experimental radar data. Results demonstrate the effectiveness of the proposed method for non-stationary signal separation and for detecting manoeuvring target in heavy sea-clutter environments. The improvement factor and clutter attenuation are calculated and used to compare the performance of the EMD and the FB-TF methods in suppressing the sea-clutter and enhancing target detection. The proposed method can be used as a potential tool for detecting and enhancing the low observable manoeuvring air targets in the sea-clutter environment.

Detection performance analysis of recurrence quantification analysis measures for low observable target within sea clutter under different sea conditions

An approach based on recurrence quantification analysis (RQA) measures is proposed to detect the low observable target within sea clutter in this study. Based on the criterion named the fixed amount of nearest neighbours, recurrence plots (RPs) are constructed. RQA measures are used to describe the line structures of the RPs. According to the differences between the RQA measures of sea clutter with and without a target, the low observable target within sea clutter can be detected. Four sea clutter datasets under different sea conditions in the IPIX OHGR database are selected to analyse the target detection performance of RQA measures. Compared with the multifractal correlation spectrum approach and the cell average constant false alarm rate detector, the author's approach can achieve a higher detection probability. Furthermore, the authors analyse the receiver operating characteristic curves of RQA measures in all of polarisations under different sea conditions. The experimental results show that the horizontal–horizontal polarisation dataset and one of the horizontal–vertical and vertical–horizontal polarisation datasets should be selected for the target detection, and the significant wave height has a greater influence on the target detection performance than the wind speed does.

Track Detection of Low Observable Targets Using a Motion Model

A method for detecting a low observable target track using an acceleration-based overall motion model is proposed. Unlike the existing track-before-detect methods that are based on sequential state updates, this method computes integrated echo energy for the entire hypothesized motion. The detection and the estimation of the track are made simultaneously using the batch processing approach. A comparison of track detection probability shows higher performance against low observable targets. Using a motion similarity metric and motion model homogeneity, a performance prediction model is derived and compared with the simulation results.

ML–PDA and ML–PMHT: Comparing Multistatic Sonar Trackers for VLO Targets Using a New Multitarget Implementation

The maximum-likelihood probabilistic data association (ML-PDA) tracker and the maximum-likelihood probabilistic multihypothesis (ML-PMHT) tracker are tested in their capacity as algorithms for very low observable (VLO) targets (meaning 6-dB postsignal processing or even less) and are then applied to five synthetic benchmark multistatic active sonar scenarios featuring multiple targets, multiple sources, and multiple receivers. Both methods end up performing well in situations where there is a single target or widely spaced targets. However, ML-PMHT has an inherent advantage over ML-PDA in that its likelihood ratio (LR) has a simple multitarget formulation, which allows it to be implemented as a true multitarget tracker. This formulation, presented here for the first time, gives ML-PMHT superior performance for instances where multiple targets are closely spaced with similar motion dynamics.

DGTD for a Class of Low-Observable Targets: A Comparison With MoM and (2, 2) FDTD

The simulation of low-observable targets requires high accuracy, both in the geometrical discretization as well as in the numerical solution of the electromagnetic problem. In this letter, we employ the well-known NASA almond to illustrate the accuracy of the leapfrog discontinuous Galerkin method, combined with a local time-stepping algorithm, comparing it to the method of moments (MoM) and the (2, 2) finite-difference time-domain (FDTD) methods.

Tracking the Tracker from its Passive Sonar ML-PDA Estimates

Target motion analysis (TMA) with wideband passive sonar has received much attention. Maximum likelihood probabilistic data association (ML-PDA) represents an asymptotically efficient estimator for deterministic target motion, and is especially well suited for low-observable targets; the results presented here apply to situations with higher signal-to-noise ratio as well, including of course the situation of a deterministic target observed via “clean” measurements without false alarms or missed detections. Here we study the inverse problem, namely, how to identify the observing platform (following a “two-leg” motion model) from the results of the target estimation process, i.e., the estimated target state and the Fisher information matrix (FIM), quantities we assume an eavesdropper might intercept. We tackle the problem and we present observability properties, with supporting simulation results.

Small moving infrared space target tracking algorithm based on probabilistic data association filter

Numerous false alarms for low signal-to-noise ratio (SNR) would seriously debase the performance for infrared low observable (LO) space target tracking. Due to the motion (i.e. azimuth, elevation and their derivative velocity), amplitude and size of infrared target are almost invariable and highly correlative, a multi-feature association approach based on probabilistic data association (PDA) is presented to track target in this paper. Firstly, the motion, amplitude and size of target are modeled as stationary random signal afforded Gaussian distribution. The probability of motion, amplitude and size of measurement originated as the target of interest is then estimated by Gaussian distribution, and that of false alarm is distributed uniformly. Subsequently, the combined probability of motion, amplitude and size is derived by PDA, and their weight coefficients are estimated adaptively according to their fluctuations. Finally, the relevant parameters including combination measurement are predicted and updated. Some experiments are included and the results show that the performance of target tracking by the proposed approach is significantly enhanced.

The Fractal Properties of Sea Clutter and Their Applications in Maritime Target Detection

This letter studies the fractal properties of sea clutter, mainly including the box-counting dimensions and multifractal spectra. Those properties of sea clutter with targets and sea clutter without targets are compared and found different, which enables us to develop methods to detect low-observable targets within sea clutter. An integral test is utilized to compare the multifractal spectra of sea clutter under different conditions. In practice, researchers can hardly get enough real data of variable signal-to-clutter ratio (SCR) to get systemic conclusions. Therefore, we construct new sea clutter series from real data with variable SCR. The results show that the detecting methods based on fractal analysis have strong performance of maritime target detection at very low SCR.

A New Feature Based on the Fluctuation Analysis to Detect Low Observable Target within Sea Clutter

In this paper, a novel approach based on the fluctuation analysis for the target detection in sea clutter is proposed. The self-affinity and scaling behaviors of sea clutter is analyzed by using the mean fluctuation. The q-th order normalized slope of the mean fluctuation curve is used as the characteristic parameter to describe the fractal property. The tests on the real data show that the target could be clearly distinguished from the sea clutter background with the proposed approach.

Integrated track initialization and maintenance in heavy clutter using probabilistic data association

Target tracking in high clutter or low signal-to-noise ratio (SNR) environments is an important topic and still a challenging task. Joint Maximum Likelihood Probabilistic Data Association (JML-PDA) is a well-known batch method for initializing the tracks of very low observable (VLO) targets in heavy clutter environments. On the other hand, the Joint Probabilistic Data Association (JPDA) algorithm, which is commonly used for recursive track maintenance, lacks track initialization capability. In this paper, we propose a Combined JML-PDA and JPDA (CJML-PDA) algorithm to automatically initialize and maintain the tracks. This combined approach seamlessly shares information between the initialization and maintenance stages of the tracker. In contrast, in other batch-recursive approaches the initialization and maintenance algorithms operate rather independent of each other. The effectiveness of the proposed algorithm is demonstrated on a heavy clutter scenario and its performance is tested on closely-spaced (but resolved) targets with association ambiguity using angle-only measurements.

Passive Range Estimation Using Two and Three Optical Cameras

In passive tracking of a flying object with bearing only measurements, range information is needed. This is accomplished by applying more than one passive optical imager. Active sensing systems have disadvantages for low-flying targets in cluttered backgrounds meanwhile are simply detected. In the observable regions of the target range, using two cameras, target range can be calculated by triangulation. If the target locates in the vicinity of sensors baseline, low observable target range will be resulted. This problem can be solved using three cameras. In this paper novel equations to calculate the target range using 3-D triangulation by two cameras are extracted, observability condition for target range is stated, target range is estimated by extended Kalman filter, observable target range is resulted by using three bearing sensors placed at a distance not along a straight line, and finally sensor fusion for estimated target range is performed.