Target Kinematics Research Articles

Robust adaptive extended target filtering based on multiplicative noise model

Abstract This study proposes two novel methods for tracking extended target with elliptical shape under non-Gaussian noise conditions.To achieve this, an explicit nonlinear measurement equation is formulated by incorporating a multiplicative noise term that connects the target's kinematic and shape parameters with the measurements.This measurement equation is employed to derive a closed-form solution for the Student's t extended Kalman filter based on multiplicative noise model (St-EKF-MNM)}with recursive measurement updates.To further optimize the target tracking performance, the Hessian matrix was used to deal with the nonlinearity in the measurement equation, further derive the Student's t second-order extended Kalman filter based on multiplicative noise model (St-SOEKF-MNM).The effectiveness and robustness of the proposed methods are demonstrated with simulation.

Extended target tracking using neural network and Gaussian process

AbstractIn extended target tracking, Gaussian Process (GP) is utilized to model unknown contour functions based on the model‐predicted target center and contour measurements. However, model prediction relies on accurate prior knowledge. When the model‐predicted target center is inaccurate, it will affect the modelling of the measurement model. To address issue, this letter introduces a hybrid‐driven approach that combines extended Kalman filter using GP with neural network; proposes an extended target tracking algorithm using neural network and GP. The algorithm predicts the target center according to the neural network and the target's kinematic model, and takes the prediction center and the contour measurements at the current moment as the input of the neural network, which in turn provides real‐time estimates for the predicted center compensation. The simulation results show that the algorithm has a significant improvement in tracking performance and better accuracy in estimating the center position and extent state of the target.

Tuning in to a hip-hop beat: Pursuit eye movements reveal processing of biological motion

Smooth pursuit eye movements are mainly driven by motion signals to achieve their goal of reducing retinal motion blur. However, they can also show anticipation of predictable movement patterns. Oculomotor predictions may rely on an internal model of the target kinematics. Most investigations on the nature of those predictions have concentrated on simple stimuli, such as a decontextualized dot. However, biological motion is one of the most important visual stimuli in regulating human interaction and its perception involves integration of form and motion across time and space. Therefore, we asked whether there is a specific contribution of an internal model of biological motion in driving pursuit eye movements. Unlike previous contributions, we exploited the cyclical nature of walking to measure eye movement's ability to track the velocity oscillations of the hip of point-light walkers. We quantified the quality of tracking by cross-correlating pursuit and hip velocity oscillations. We found a robust correlation between signals, even along the horizontal dimension, where changes in velocity during the stepping cycle are very subtle. The inversion of the walker and the presentation of the hip-dot without context incurred the same additional phase lag along the horizontal dimension. These findings support the view that information beyond the hip-dot contributes to the prediction of hip kinematics that controls pursuit. We also found a smaller phase lag in inverted walkers for pursuit along the vertical dimension compared to upright walkers, indicating that inversion does not simply reduce prediction. We suggest that pursuit eye movements reflect the visual processing of biological motion and as such could provide an implicit measure of higher-level visual function.

Physics-Aware Generative Adversarial Networks for Radar-Based Human Activity Recognition

Generative adversarial networks (GANs) have recently been proposed for the synthesis of RF micro-Doppler signatures to address the issue of low sample support and enable the training of deeper neural networks (DNNs) for enhanced RF signal classification. But GANs suffer from systemic kinematic inconsistencies that decrease performance when GAN-synthesized data is used for training DNNs in human activity recognition. As a solution to this problem, this paper proposes the design of a multi-branch GAN (MBGAN), which integrates domain knowledge into its architecture, and physics-aware metrics based on correlation and curve-matching in the loss function. The quality of the synthetic samples generated is evaluated via image quality metrics, the ability to synthesize data that reflects human physical properties and generalize to broader subject profiles, and the achieved classification accuracy. Our experimental results show the proposed approach generates synthetic data for training that more accurately matches target kinematics, resulting in an increase of 9% in classification accuracy when classifying 14 different ambulatory human activities.

Spherical Earth Based Measurement Modeling for Practical OTHR Target Tracking

We revisit the problem of modeling of the over-the-horizon radar (OTHR) target tracking. The accuracy of target kinematic and measurement models are critical factors affecting tracking performance. In most of the existing literature, a flat Earth OTHR measurement model is used. This has two obvious drawbacks: (1) a state vector with components of ground range, range-rate, bearing, and bearing-rate cannot be used directly in common kinematic models such as the nearly constant velocity (NCV), nearly constant acceleration (NCA), nearly constant turn (NCT), and a combination of these. (2) The curvature of Earth is ignored, which affects the measurement accuracy of the OTHR. We develop a spherical Earth based measurement model. First, the target kinematic state is modeled in a two-dimensional tracker coordinate frame (T frame), which is suitable for the common kinematic models. Second, the measurement modeling is completed using the spherical Earth model, which fully considers the curvature of Earth among the receiver, transmitter, and target. Furthermore, the inverse mapping for track initiation is derived. Finally, the effectiveness of the developed model is demonstrated by a non-maneuvering target moving with the NCV motion and a maneuvering target moving with a sequence of NCV, NCT, and NCV motions. The interacting multiple model (IMM) algorithm is used for state estimation of the maneuvering target. The unscented Kalman filter (UKF), cubature Kalman filter (CKF), and quadrature Kalman filter (QKF) are compared for state estimation.

SAR Signal Formation and Image Reconstruction of a Moving Sea Target

Maritime application of synthetic aperture radar (SAR) technology for sea-target surveillance and imaging is considered in this study. A SAR scenario, including the kinematics of a SAR satellite and a ship moving on the sea, along with the geometry of the target, are analytically described. A linear frequency modulation (LFM) waveform is applied for the target’s illumination. Based on the target’s geometry, SAR and target kinematics and the LFM waveform, a SAR signal model is synthesized. It is proven that the process of signal formation is a transformation of the three-dimensional (3D) image into a two-dimensional (2D) signal, whereas the target’s 2D imaging is an inverse transformation of the 2D signal into the target’s 2D image. SAR signal components, linear Fourier terms and higher-order phase terms are analytically derived and discussed in detail. Moreover, it is proven that SAR image reconstruction is a motion-compensation procedure, i.e., it removes all phases induced by first- and higher-order motion. Based on the SAR signal analysis, an illustrative iterative image-reconstruction algorithm is derived. The quality of the imaging is evaluated by an entropy cost function. Simulation experiments are carried out to verify the correctness of the theoretical statements in respect of SAR signal formation and image reconstruction.

The KOALA experiment for (anti)proton–proton elastic scattering

The KOALA experiment measures the differential cross section of (anti)proton–proton elastic scattering over a wide range of four-momentum transfer squared |t| from 0.0008 to 0.1 (GeV/c)2. These data are essential input for the PANDA experiment to reach the desired absolute precision for the integrated luminosity determination. The KOALA experiment is based on fixed target kinematics and uses an internal hydrogen cluster jet target. The wide range of |t| is achieved by measuring the total kinetic energy of the recoil protons near 90°. The recoil detector consists of silicon and germanium single-sided strip sensors with an energy resolution of approximately 20keV and 30keV (FWHM), respectively. A forward detector consisting of two layers of plastic scintillators measures the elastically scattered beam particles in the forward direction close to the beam axis. By requiring a coincidence of the forward detector with the recoil detector, the large background at small recoil angles is suppressed, thereby improving the identification of elastic scattering events in the low |t| range. The KOALA setup has been installed and commissioned with proton beam at COSY in order to validate the detector by measuring the proton–proton elastic scattering. The results from this commissioning are presented here.

Autonomous angles-only multitarget tracking for spacecraft swarms

This paper presents a new algorithm for autonomous multitarget tracking of resident space objects using optical angles-only measurements from a spaceborne observer. To enable autonomous angles-only navigation of spacecraft swarms, observers must identify and track multiple known or unknown target space objects in view, without reliance on a-priori relative orbit knowledge. Extremely high tracking precision is necessary despite low measurement frequencies and limited computational resources. The new ’Spacecraft Angles-only MUltitarget tracking System’ (SAMUS) algorithm has been developed to meet these objectives and constraints. It combines domain-specific modeling of target kinematics with multi-hypothesis techniques to autonomously track multiple unknown targets using only sequential camera images. A measurement transform ensures that target motion in the observer reference frame follows a consistent parametric model; curve fitting is used to predict track behavior; and kinematic track gating and scoring criteria improve the efficiency and accuracy of the multi-hypothesis approach. Monte Carlo testing with high-fidelity simulations demonstrates close to 100% data association precision and high recall across a range of multi-spacecraft formations, in both near-circular and eccentric orbits. Tracking is maintained in the presence of eclipse periods, significant measurement noise, and partially known swarm maneuvers. A comparison to other tracking algorithms reveals strong advantages in precision, robustness and computation time, crucial for spaceborne angles-only navigation.

Experimental Demonstration of the Lower Leg Trajectory Error Framework Using Physiological Data as Inputs.

While many studies have attempted to characterize the mechanical behavior of passive prosthetic feet to understand their influence on amputee gait, the relationship between mechanical design and biomechanical performance has not yet been fully articulated from a fundamental physics perspective. A novel framework, called lower leg trajectory error (LLTE) framework, presents a means of quantitatively optimizing the constitutive model of prosthetic feet to match a reference kinematic and kinetic dataset. This framework can be used to predict the required stiffness and geometry of a prosthesis to yield a desired biomechanical response. A passive prototype foot with adjustable ankle stiffness was tested by a unilateral transtibial amputee to evaluate this framework. The foot condition with LLTE-optimal ankle stiffness enabled the user to replicate the physiological target dataset within 16% root-mean-square (RMS) error. Specifically, the measured kinematic variables matched the target kinematics within 4% RMS error. Testing a range of ankle stiffness conditions from 1.5 to 24.4 N·m/deg with the same user indicated that conditions with lower LLTE values deviated the least from the target kinematic data. Across all conditions, the framework predicted the horizontal/vertical position, and angular orientation of the lower leg during midstance within 1.0 cm, 0.3 cm, and 1.5 deg, respectively. This initial testing suggests that prosthetic feet designed with low LLTE values could offer benefits to users. The LLTE framework is agnostic to specific foot designs and kinematic/kinetic user targets, and could be used to design and customize prosthetic feet.

Do deep networks encode a similar representation compared to a model of target kinematics during an interception task?

Do deep networks encode a similar representation compared to a model of target kinematics during an interception task?

Dynamic model‐based feature aided data association filter in target tracking

In this study, the authors consider the data association problem where the feature of a target is available. Most existing data association filters mainly use a statistical or simple model of the feature without explicitly considering the correlation between the target behaviour and feature characteristics. The inaccurate model of the feature could lead to divergence of the estimation error or the loss of a target in heavily cluttered and/or low signal-to-noise ratio environments. To address the problem, they first develop feature models (e.g. target dimensions) dependent on the target behaviour (i.e. the distance between the sensor and the target, and the aspect angle between the longitudinal axis of the target and the axis of sensor line of sight). Then they propose a data association filter which can facilitate the feature models dependent on the target kinematics to reduce the misassociations. The performance of the proposed feature models and data association filter are demonstrated with illustrative target tracking scenarios.

Machine learning-based classification of recreational fishing vessel kinematics from broadband striation patterns

Machine learning is applied to the classification of underwater noise for rapid identification of surface vessel opening and closing behavior. The classification feature employed is the broadband striation pattern observed in a vessel's acoustic spectrogram measured at a nearby hydrophone. Convolutional neural networks are particularly well-suited to the recognition of textures such as interference patterns in broadband noise radiated from moving vessels. Such patterns are known to encode information related to the motion of its source. Rapid understanding of target kinematics through machine learning can provide powerful and informative cues as to the identity and behavior of a detected surface vessel.

Effect of walking with a modified gait on activation patterns of the knee spanning muscles in people with medial knee osteoarthritis

ObjectiveTo evaluate muscle activation patterns and co-contraction around the knee in response to walking with modified gait patterns in patients with medial compartment knee-osteoarthritis (KOA). Design40 medial KOA patients walked on an instrumented treadmill. Surface EMG activity from seven knee-spanning muscles (gastrocnemius, hamstrings, quadriceps), kinematics, and ground reaction forces were recorded. Patients received real-time visual feedback on target kinematics to modify their gait pattern towards three different gait modifications: Toe-in, Wider steps, Medial Thrust. The individualized feedback aimed to reduce their first peak knee adduction moment (KAM) by ≥10%. Changes in muscle activations and medial/lateral co-contraction index during the loading response phase (10–35% of the gait cycle) were evaluated, for the steps in which ≥10% KAM reduction was achieved. ResultsData from 30 patients were included in the analyses; i.e. all who could successfully reduce their KAM in a sufficient number of steps by ≥10%. When walking with ≥10% KAM reduction, Medial Thrust gait (KAM −31%) showed increased flexor activation (24%), co-contraction (17%) and knee flexion moment (35%). Isolated wider-step gait also reduced the KAM (−26%), but to a smaller extent, but without increasing muscle activation amplitudes and co-contraction. Toe-in gait showed the greatest reduction in the KAM (−35%), but was accompanied by an increased flexor activation of 42% and hence an increased co-contraction index. ConclusionGait modifications that are most effective in reducing the KAM also yield an increase in co-contraction, thereby compromising at least part of the effects on net knee load.

Partial integrated guidance and control design for supersonic missile based on disturbance rejection

In this paper, a three-dimensional partial integrated guidance and control law for the supersonic missile impacting the surface target is proposed. In the guidance loop design, a reduced-order extended state observer is integrated with feedback linearization and the pure proportional navigation law to eliminate the effects of the unknown target kinematics and the dynamic coupling between the vertical and horizontal planes. Considering the autopilot lag, two guidance schemes with different observer inputs are compared using the adjoint and frequency domain analyses. For the control loop, the angular velocity feedback is performed first for the pitch and yaw channels as conventionally done in practice, and then first-order linear active disturbance rejection control is designed for the damping-enhanced plant to deal with the fast time-varying, uncertain aerodynamics and the strong coupling effects among the three channels. Both the attitude angle and the acceleration control can be designed in a unified framework. Finally, a three-dimensional engagement scenario is performed using a six-degree-of-freedom supersonic vehicle model, and extensive simulations are carried out to validate the robustness and the high guidance precision of the proposed method.

Pressure Distribution and Shear Forces inside the Proton.

The distributions of pressure and shear forces inside the proton are investigated using lattice quantum chromodynamics (LQCD) calculations of the energy momentum tensor, allowing the first model-independent determination of these fundamental aspects of proton structure. This is achieved by combining recent LQCD results for the gluon contributions to the energy momentum tensor with earlier calculations of the quark contributions. The utility of LQCD calculations in exploring, and supplementing, the assumptions in a recent extraction of the pressure distribution in the proton from deeply virtual Compton scattering is also discussed. Based on this study, the target kinematics for experiments aiming to determine the pressure and shear distributions with greater precision at Thomas Jefferson National Accelerator Facility and a future electron ion collider are investigated.

Relationship Between Finite Set Statistics and the Multiple Hypothesis Tracker

The multiple hypothesis tracker (MHT) and finite set statistics (FISST) are two approaches to multitarget tracking, which both have been heralded as optimal. In this paper, we show that the multitarget Bayes filter with basis in FISST can be expressed in terms the MHT formalism, consisting of association hypotheses with corresponding probabilities and hypothesis-conditional densities of the targets. Furthermore, we show that the resulting MHT-like method under appropriate assumptions (Poisson clutter and birth models, no target death, linear-Gaussian Markov target kinematics) only differs from Reid's MHT with regard to the birth process.

A Novel Framework for Quantitatively Connecting the Mechanical Design of Passive Prosthetic Feet to Lower Leg Trajectory.

This paper presents a novel framework that quantitatively connects the mechanical design of a prosthetic foot to its anticipated biomechanical performance. The framework uses kinetic inputs (ground reaction forces and center of pressure) to predict kinematic outputs of the lower leg segment by knowing the geometry and stiffness of the foot. The error between the predicted and target kinematics is evaluated using a root-mean-square error function called the Lower Leg Trajectory Error (LLTE). Using physiological kinetic inputs and kinematic targets, three model foot architectures were optimized to minimize the LLTE. The resulting predicted lower leg kinematics were compared to those of the same foot architectures optimized for physiological roll-over geometry. The feet with minimized LLTE had lower leg kinematics closer to physiological than those optimized for roll-over geometry. A prosthetic foot that exactly mimics physiological roll-over geometry may result in gait kinematics that differ greatly from physiological, as roll-over geometry omits information about the foot-ground contact constraint, lower leg orientation, and temporal progression of a step. The LLTE-based framework is agnostic to specific foot designs provided their constitutive behavior can be characterized, and it can accept alternate inputs and targets depending on what performance and clinical objectives are desired.

An ellipse extended target CBMeMBer filter using gamma and box-particle implementation

This work presents a gamma box-particle implementation of the ellipse extended target cardinality balanced multi-target multi-Bernoulli (ET-CBMeMBer) filter, named the ellipse extended target gamma box particle CBMeMBer (EET-GBP-CBMeMBer) filter, where the ellipse extended targets are characterized using a Poisson model suggested by Gilholm et al. The proposed filter is capable of dynamically tracking multiple ellipse extended targets and estimating the target kinematics and extents as well as the measurement rates of targets, in the presence of clutter measurements, false alarms and missed detections. To get the CBMeMBer recursion of the EET-GBP-CBMeMBer filter, a suitable measurement likelihood is derived for a given partitioning cell, and the main gamma box-particle implementation are given along with the necessary box manipulations and approximations. The capabilities and limitations of the proposed EET-GBP-CBMeMBer filter are demonstrated by simulation examples. The simulation results show that a gamma box-particle implementation of the ET-CBMeMBer filter can avoid the high number of particles and reduce computational burden, compared to a particle implementation of that for extended target tracking.

Fault-tolerant feature-based estimation of space debris rotational motion during active removal missions

One of the key functionalities required by an Active Debris Removal mission is the assessment of the target kinematics and inertial properties. Passive sensors, such as stereo cameras, are often included in the onboard instrumentation of a chaser spacecraft for capturing sequential photographs and for tracking features of the target surface. A plenty of methods, based on Kalman filtering, are available for the estimation of the target's state from feature positions; however, to guarantee the filter convergence, they typically require continuity of measurements and the capability of tracking a fixed set of pre-defined features of the object. These requirements clash with the actual tracking conditions: failures in feature detection often occur and the assumption of having some a-priori knowledge about the shape of the target could be restrictive in certain cases. The aim of the presented work is to propose a fault-tolerant alternative method for estimating the angular velocity and the relative magnitudes of the principal moments of inertia of the target. Raw data regarding the positions of the tracked features are processed to evaluate corrupted values of a 3-dimentional parameter which entirely describes the finite screw motion of the debris and which primarily is invariant on the particular set of considered features of the object. Missing values of the parameter are completely restored exploiting the typical periodicity of the rotational motion of an uncontrolled satellite: compressed sensing techniques, typically adopted for recovering images or for prognostic applications, are herein used in a completely original fashion for retrieving a kinematic signal that appears sparse in the frequency domain. Due to its invariance about the features, no assumptions are needed about the target's shape and continuity of the tracking. The obtained signal is useful for the indirect evaluation of an attitude signal that feeds an unscented Kalman filter for the estimation of the global rotational state of the target. The results of the computer simulations showed a good robustness of the method and its potential applicability for general motion conditions of the target.

Eye-hand coordination during flexible manual interception of an abruptly appearing, moving target.

As a vital skill in an evolving world, interception of moving objects relies on accurate prediction of target motion. In natural circumstances, active gaze shifts often accompany hand movements when exploring targets of interest, but how eye and hand movements are coordinated during manual interception and their dependence on visual prediction remain unclear. Here, we trained gaze-unrestrained monkeys to manually intercept targets appearing at random locations and circularly moving with random speeds. We found that well-trained animals were able to intercept the targets with adequate compensation for both sensory transmission and motor delays. Before interception, the animals' gaze followed the targets with adequate compensation for the sensory delay, but not for extra target displacement during the eye movements. Both hand and eye movements were modulated by target kinematics, and their reaction times were correlated. Moreover, retinal errors and reaching errors were correlated across different stages of reach execution. Our results reveal eye-hand coordination during manual interception, yet the eye and hand movements may show different levels of prediction based on the task context. NEW & NOTEWORTHY Here we studied the eye-hand coordination of monkeys during flexible manual interception of a moving target. Eye movements were untrained and not explicitly associated with reward. We found that the initial saccades toward the moving target adequately compensated for sensory transmission delays, but not for extra target displacement, whereas the reaching arm movements fully compensated for sensorimotor delays, suggesting that the mode of eye-hand coordination strongly depends on behavioral context.