Search

In this paper, we propose a novel probabilistic track initiation algorithm using velocity information of measurements. The velocity information is converted to a probabilistic function and it is used to calculate the track score for track initiation. Moreover, we introduce the scheme of Non-Maximum Suppression (NMS) to determine candidate targets for tracking in terms of the track score. To verify the effectiveness of the proposed algorithm, we designed a simulator for target tracking. The simulation results show that the proposed algorithm has lower false track probability compared to conventional algorithms in heavy clutter environments.

The probability hypothesis density (PHD) filter has been proved to be an effective method for multi-target tracking in the presence of false alarms, missed detections and an unknown numbers. However, the PHD filter fails to obtain the state estimation of a target accurately during its initial time. This will degrade the performance of the PHD filter on the detection of new targets. In order to resolve this problem, we apply the rule-based track initiation technique to the Gaussian mixture PHD filter, and propose the Gaussian mixture PHD filter with track initiation. The results of the simulation experiment demonstrate that the proposed filter can achieve better performance on tracking multi-target than the Gaussian mixture PHD filter.

This talk goes through the components in generic single-scan target tracking algorithms from filtering to data association, track initiation, and termination. In many areas, reference is made to functions in the open-source copyleft-free Tracker Component Library (available online) so that attendees can rapidly apply the algorithms that are discussed.

International audience; In this paper, we investigate a new track initiation technique enabling the use of a low-cost radar system for Low-Earth-Orbit surveillance. This technique is based on a first association of observations with little ambiguity followed by a fast Initial Orbit Determination. This study supports the feasibility of the system as this technique shows a coverage of 84,4% within 6 days, with a combinatorial complexity kept under control when assessed in a realistic multitarget tracking context.

A neutrosophic Hough transform-based track initiation method (NHT-TI) is proposed to solve the uncertain track initiation problem in a complex surveillance environment. In the proposed method, a neutrosophic set is employed to describe the uncertain association of a measurement with different targets, which is divided into three categories, including the association with real targets, uncertain targets, and false targets, respectively. On this basis, the neutrosophic Hough transform (NHT) method is developed in the framework of the standard Hough transform (HT) method. Based on the sequential processing mode of the sensors, candidate temporary tracks are further defined to directly calculate the parameter points in the parameter space, which are utilized to calculate the contribution for the corresponding vote cells in the accumulation matrix by using the Gaussian membership function. Based on the scheme, the NHT method can avoid large traversal operations and reduce computation complexity of the HT method. Moreover, considering the effects of noises and clutters on track initiation, two constraint conditions related to the velocity information of moving targets and the time information of measurement sequences are introduced to suppress false tracks and reduce vote times. Finally, the real tracks can be determined by detecting the peaks of the global accumulation matrix. The performance of the proposed NHT-TI method is evaluated by using two experiments with simulated data and real data in two complex surveillance environments. The results are found to be better than those of the standard HT-based track initiation (HT-TI) method, the modified HT-TI method based on candidate temporary tracks and the improved HT-TI method in detection reliability and computational complexity.

This paper addresses the problem of track initiation for multi-target in different motion forms and in complicated clutter background. The method proposed combines the traditional logic-based method and convolution neural network. The logic-based method is used mainly to generate a set of track proposals, which is computed by the convolution neural network to extract features in data domain. In this paper, softmax at the end of the convolution neural network is substituted by a one-dimensional two-class classifier for the output layer of the convolution neural network is designed to output a one-dimensional value. There are two key insights in this method: (1) the classification problem has been transformed into target tracking problem on the condition that the set of track proposals is found. (2) the convolution neural network is firstly used in data domain to mine and augment high-level features that make classification more easily. The simulation experiments have shown that this method performs much better than modified Hough transform which is used to initialize tracks traditionally, especially when the targets are maneuver. In the experiments based on real data, this method is proved to be adaptive enough to initialize tracks whose data comes from different radars.

The detection and tracking of an unknown number of targets using a Bayesian hierarchical model with target labels is presented. To approximate the posterior probability density function (PDF), we develop a two-layer particle filter (PF). One deals with track initiation, and the other deals with track maintenance. In addition the parallel partition (PP) method is proposed to sample the states of the surviving targets.

Part 2: Data Mining; International audience; This paper deals with the problem of track initiation in dense clutters. A novel Hough transform method based on prior motion information is proposed. Firstly, with time sequence of the measurements and targets’ kinematic information considered, the single scan data accumulation effect is avoided, and the initiated tracks are more reliable. Then, measurements in observation space only vote to part of the lines that pass through them, which can effectively reduce the computational cost, and improve the initiation speed. Simulation results indicate that the proposed method is superior to the traditional algorithms in terms of detection probability, computational cost and robustness.

Purely from the perspective of target tracking, we have completed the pretreatment of group-target measurement, group-target track initiation, single group-target data correlating and track maintenance, and multi-group-target data correlating and track maintenance.

Problems and solutions are discussed for the application of the Probability Hypothesis Density (PHD) filter in radar system. The results for engineering applications of the multiple target tracking based on PHD filtering in radar system are introduced. Results showed that the PHD filter can effectively solve the problems of Track initiation, organizing batch, and so on,.

This paper firstly proposed a universal track initiation method with adaptive threshold. Using this method, performance under heavy clutter tracking environment is improved. Besides, joint probabilistic data association filter is the common data association method, which only deal with the fixed-numbered multi-target tracking problem. Introducing the data pre-association process, an improved JPDAF algorithm is proposed to accomplish variable-numbered multi-target tracking, which is validated by simulation results.

This paper presents a track-before-detect (TBD) algorithm for detection of unknown quantity of targets with parabolic tracks. First, eliminate large number of clutters orderly by expanded trellis and the method of mean power. And then get candidate parabolas by Randomized Hough Transform (RHT). The true tracks of the targets are extracted successfully by a strategy of outliers eliminating at last. Experimental results indicate that the proposed algorithm is capable of detecting multiple targets without the assumptions of an upper bound to the number of targets.

A fundamental characteristic of PMHT (Probabilistic Multi-Hypothesis Tracker) is that the number of targets and initial states of targets in the surveillance area must be a priori known. This requirement is impossible to fulfil in almost every realistic scenario. In the paper, we present two track initiation methods to solve the problem. The proposed track initiation methods are 2-point track initiation and Hough transform track initiation, and they are used to evaluate track initial states and weights for FTD (False Track Discrimination) of the PMHT algorithm. Also suggested as automatic target detection for tracking systems that combines track initiation for target detection with the PMHT algorithm for target tracking in a cluttered environment. A series of Monte-Carlo simulation runs is employed to evaluate the overall system performance with the two track initiation methods and the results are compared and analyzed.

This article proposes algorithms for the automatic initiation of the tracks of maneuvering targets in cluttered environments. These track initiation algorithms consist of IPDA-AI(Integrated Probabilistic Data Association-Amplitude Information) and MPDA(Most Probable Data Association) in an Interacting Multiple Model(IMM) configuration, and they are referred to as the IMM-IPDAF-AI and IMM-MPDA respectively. The IMM portion consists of several filters based on different dynamical models to handle target maneuvers. Each of the filters utilizes an IPDA-AI(or MPDA) algorithm to deal with the problem of track existence in the presence of clutter. Although the primary purpose of this study is to deal with the track initiation problem, the IMM-IPDAF-AI and IMM-MPDA can also be used for the maintenance of existing tracks and the termination of tracks for targets when they disappear. For illustrative purposes, simulation is used to compare the performance of the algorithms proposed to other track formation algorithms.

Probability-1 homotopy algorithms are seldom taught in standard numerical analysis classes, and, with a few exceptions, their utility in target track initiation has been largely over-looked. This paper demonstrates how such methods can be used for unbiased measurement conversion and initial target-state estimation in difficult scenarios involving non-simultaneous measurements when combined with cubature integration techniques. Scenarios considered are range-only and time-delay-of-arrival-only measurements as well as surface-wave time-delay-of-arrival-only measurements.

In this paper, we present an online approach for joint initiation/termination and tracking for multiple targets with multiple sensors using sequential Monte Carlo (SMC) methods. There are several main contributions in the paper. The first contribution is the extension of the deterministic initiation and termination method proposed by the authors' previous publications to a full SMC context in which track initiation/termination are executed with sampling methods. In effect, the dimensions of the particles are variable. In addition, we also integrate a Markov random field (MRF) motion model with the framework to enable efficient and accurate tracking for interacting targets and to avoid potential track coalescence problems. With the employment of multiple sensors, a centralized tracking strategy is adopted, where the observations from all active sensors are fused together for target initiation/termination and tracking and a set of global tracks is maintained. Intra-and inter-sensor clusters are constructed, comprised of closely spaced observations either in time for single sensors or from distinct sensors at a single time, that can increase the reliability when proposing new tracks for initiation. Computer simulations demonstrate that the proposed approach is robust in joint initiation/termination and tracking of multiple manoeuvring targets even when the environment is hostile with high-clutter rates and low target detection probabilities. The integration of the MRF framework into the proposed methods improves robustness in handling close target interactions when the observation noise is high.

A multiextended-target tracker based on the extended target Gaussian-mixture probability hypothesis density (ET-GMPHD) filter, which can provide the tracks of the extended targets, is proposed to maintain the track continuity for the extended targets. To identify the extended targets, each individual Gaussian term of the mixture representing the posterior intensity function will be assigned a label, which is evolved through time. Then a track management scheme, including track initiation, track confirmation, track propagation, and termination, is developed to form the tracks for the extended targets. Furthermore, to improve the performance of the extended target tracker we also propose a mixture partitioning algorithm for resolving the identities of the extended targets in close proximity. The simulation results show that our proposed tracker achieves the less error of the position estimates and decreases the probability of incorrect label assignments from 0.6 to 0.25.

Theoretical design of a tracker with respect to the so-called target perceivability is presented. Basic rules and analytic formulas for the determination of tracker parameters are presented for making better tracking decisions (i.e., track initiation, confirmation, and termination) toward enhancing tracking performance. Simulation results are provided that support the theoretical design and demonstrate the enhancement of the tracker in tracking performance.

The paper analyses the improvement of track loss in clutter with multisensor data fusion. By a determination of the transition probability density function for the fusion prediction error, one can study the mechanism of track loss analytically. With nearest-neighbor association algorithm. The paper we studies the fused tracking performance parameters, such as mean time to lose fused track and the cumulative probability of lost fused track versus the normalized clutter density, for track continuation and track initiation, respectively. A comparison of the results obtained with the case of a single sensor is presented. These results show that the fused tracks of multisensor reduce the possibility of track loss and improve the tracking performance. The analysis is of great importance for further understanding the action of data fusion.

This paper proposes a novel track initiation algorithm for the passive sensor network system which sample the measurements asynchronously and aperiodicly. Firstly, the algorithm calculates the heading-angle using the azimuths and the elevations measured by the passive sensor. And then, the candidate trajectory sets are established trough the synoptic data association. Finally, the algorithm predicts the target the azimuth and the elevation, and uses the conjugate gradient method (CGM) to identify the target state vector including the positions. We have implemented simulation experiments to evaluate this algorithm's performance. Their results show that it is appropriate for engineering applications.

Performance analysis for the track initiation phase in very high target density environments is described. The analysis of tracking performance performed by S. Mori et al. (1989, 1990) yielded very simple expressions to predict average track purity in terms of tracking parameters. These results were, however, obtained mainly for the track continuation phase in which tracks are more or less well established. The authors extend these results to the track initiation phase. A linear approximation of the method which significantly simplifies the calculation is also presented. A simple test scenario, consisting of a threat object group and an orbiting sensor, was used to demonstrate the performance evaluation for track initiation. >

Track initiation using multiple bistatic range and range rate measurements with multidimensional assignment algorithm was investigated in this work. The assignment algorithm was tested on a problem involving three targets moving in close proximity inside a two dimensional radar network of six bistatic radar pairs. The accuracy and consistency of initiated tracks as well as measurement association performance of the assignment algorithm was demonstrated through Monte Carlo simulations.

An analytical model is provided which describes the tracking process that evolves from a sequence of detection opportunities. We assume a sliding window of length N, with M (M ≤ N) detections required for track acquisition, and L consecutive misses required for loss of track. We provide a recursive solution for the probabilities of acquisition, loss, and being in track, and then define the distributions of the times of acquisition, loss, and duration for each track and gap that can develop. There are no restrictions on M, N, L, or on a varying sequence of associated detection probabilities. Numerical results are provided for typical cases.

This study investigates the effect of incorporating Doppler velocity measurements into the most commonly used sequential track initiation schemes, the rule and logic based schemes. The measurement set has been expanded from range and azimuth to include elevation and Doppler velocity. Unlike previous studies, elevation and Doppler measurements have also been included in the analytical evaluation of the false track initiation probability. Performance improvement obtained by employing Doppler measurements has been demonstrated through simulations in terms of false track initiation probability and true track initiation probability. Receiver Operating Characteristics and System Operating Characteristics have also been utilized in performance evaluation. Analytical derivations and simulations have revealed that using Doppler velocity measurements along with 3D position measurements while initiating tracks in clutter leads to significant decrease in false track initiation probability while providing an acceptable level of true track initiation probability. Simulation results have also shown that inclusion of Doppler measurements has reduced the required time for track initiation, resulting in faster track initiation.

This chapter treats the various types of radar searches, such as volume search, horizon fence search, cued search, and sector search, the different types of waveforms used, and the acquisition function used as a prerequisite for track initiation (TI). Topics include: types of searches; search design; search waveforms and processing.

This paper introduces a novel track initiation technique for the bearings-only two-sensor-multi-target tracking system. Observing that each target is of the characteristic of uniform rectilinear motion, a new objective functional derived from the thought of Hough transform is investigated. On the basis of it, the ant colony optimization (ACO) algorithm, a kind of heuristic optimization method, is utilized to initiate tracks of interest. Numerical simulation results are presented and show that the proposed ACO-based track initiation method performs well compared with other traditional ones such as rule-based, logic-based, and Hough-transform-based track initiation approaches.

Scattering of signals from commercial television or radio channels from air and surface targets can be detected by one or more passive receivers. The received signals can be processed to give bistatic range and range rate measurements on targets. This paper considers tracking algorithms to convert the measurements to target tracks using data from a demonstration receiver, with emphasis on the problem of track initiation for low SNR targets where there are many false detections. It is shown that it is possible to track targets in this type of environment. Further development and tuning is required to minimise the number of false tracks initiated.

To investigate the problem of track initiation for multi-static radar system, an algorithm with measurement association discriminants is derived based on two dimensional measurements. The S-D assignment algorithm is adopted to search minimum cost associations. Simulation results show that the algorithm can accurately calculate multiple traces with fast speed. Meanwhile, it has good stability and convergence.

This paper will discuss closely integrating the tracker and radar resource management processes within phased array radar. Several techniques are discussed and, where possible, basic analyses and figures of merit are given. In particular an analysis of time-agile track initiation in land clutter suggests significant benefits for slow moving targets. (4 pages)

Due to some adverse factors, tracking multitarget in dense clutter becomes a difficult task. Therefore, a new algorithm is presented in this paper which consists of m-best 2-D assignment algorithm and generalized likelihood ratio test (GLRT). Simulated numerical examples are given, which demonstrate that this algorithm can solve the problem of "measurement-to-track" association efficiently, and discriminate the true or false track in short time which can reduce the need for storage and calculation enormously.

Date of Conference: 7-8 March 2006 Conference Name: IEE Seminar on Target Tracking: Algorithms and Applications, 2006 Performance of multiple target tracking algorithms in complex environments heavily relies on the success of track initiation and measurement-to-track association algorithms. Doppler velocity measurement is the major discriminant of clutter from the target of interest with relatively higher velocities. This work summarizes the analytical derivations and presents simulation results about track initiation and maintenance using Doppler velocity reports along with the 3D position measurements extracted by a phased array radar.

In this paper, a Hough Transform (HT) based track initiation method is proposed to detect targets in 3L (low signal-to-clutter-ratio, low signal-to-noise-ratio, and low detection probability) environment using a new and effective accumulation method. In the new accumulation method, the contributions of each cell's votes are determined by the neighbors of the resolving cell, instead of just the cell itself so that the disturbance of sampled probability density, due to finite samples, can be smoothed. Simulation results show that our method not only makes significant improvement in reducing false tracks but also has stronger robustness to measurement noise and clutters, compared with standard HT based track initiation with binary accumulation.

Su Feng presented the modified logical track initiation algorithm in his paper, however, the development of the algorithm has not been proved in theory. In this paper, the track detection probability and false track probability of the algorithm is deduced by statistics methods. The simulation test verifies the results.

Automatic track initiation in clutter within the smoothing framework is considered in this paper. We derive a new smoothing algorithm based on integrated probabilistic data association (IPDA) filter and augmented state smoothing framework. We demonstrate the benefits of smoothing based approaches in multi-target tracking in clutter scenarios using tracker performance measures like true/false track discrimination statistics, track initiation/termination times and RMS errors of confirmed true tracks.

The fundamental problem of target tracking in a dense multiple-target environment with clutter, false alarms, missed detections, and so forth is the uncertain origin of measurements. In such a dense multiple-target environment, there is a need to perform track-initiation procedure by which the track of each target is detected. After we perform the track-initiation procedure, we can perform track-maintenance procedure. In this paper, we propose a new track-initiation method using measurements obtained by the optical sensors in a multiple-target environment. Moreover, we estimate the performance of the proposed method.

An algorithm capable of integrating the multisensor integration (MSI) track files from internetted friendly aircraft has been developed and validated through simulation. In addition to enhancing the situation awareness, the algorithm allows target position to be determined from passive sensor measurements to minimize the exposure of friendly aircraft. The position and velocity of multiple targets are estimated by subsequentially processing the angular information in MSI track files. Range measurements are not required for this process, although availability of range information, the ITFM (internetted track file maintenance) algorithm relies on passive ranging and deghosting for track initiation. >