Multi-sensor Multi-target Tracking Research Articles

Gray track-to-track correlation algorithm for distributed multitarget tracking system

Track correlation is a fundamental problem in distributed multitarget multisensor tracking system. It entails selecting the most probable association between target tracks from a very large set of ...

Multisensor target tracking performance with bias compensation

In this paper, multisensor-multitarget tracking performance with bias estimation and compensation is investigated when only moving targets of opportunity are available. First, we discuss the tracking performance improvement with bias estimation and compensation for synchronous biased sensors, and then a novel bias estimation method is proposed for asynchronous sensors with time-varying biases. The performance analysis and simulations show that asynchronous sensors have a slightly degraded performance compared with the "equivalent" synchronous ones. The bias estimates as well as the corresponding Cramer-Rao lower bound (CRLB) on the covariance of the bias estimates, i.e., the quantification of the available information on the sensor biases in any scenario are also given. Tracking performance evaluations with different sources of biases-offset biases, scale biases, and sensor location uncertainties, are also presented and we show that tracking performance is significantly improved with bias estimation and compensation compared with the target tracking using the original biased measurements. The performance is also close to the lower bound obtained in the absence of biases

Gray track-to-track correlation algorithm for distributed multitarget tracking system

Track correlation is a fundamental problem in distributed multitarget multisensor tracking system. It entails selecting the most probable association between target tracks from a very large set of possibilities. First, the inherent disadvantages of the present two kinds of track correlation algorithms are pointed out in this paper. Then a novel track correlation algorithm based on the gray theory is presented and the detailed steps are described. This method can weigh with the degree of relative correlation through the developing trend between the factors in data arrays, which breaks through the limitations of the sample quantities and the typical distributing rule. Effect of discrimination coefficient is also discussed. At last, the correct correlation rate and computational time cost of the proposed algorithm are compared with weighted algorithm and statistical double threshold algorithm through simulation. The simulation results show that the performance of gray track correlation algorithm proposed in this paper is much better than that of the classical methods under the environments of dense targets, interfering, noise, tracks cross and so on.

Tracking of Multiple Moving Speakers With Multiple Microphone Arrays

In this paper we use multiple microphone arrays to fuse the location estimate from each microphone array which yields an improved estimate of the positions and velocities of multiple, simultaneously active, moving speakers based on time delay of arrivals (TDOAs). Our approach: 1) incorporates kinematic information of moving speakers by using an interacting multiple model (IMM) estimator for each speaker in order to constrain the evolution of the location measurements; 2) fuses the location estimates of the same speaker from multiple microphone arrays for better acoustical coverage of the sensed environment, and 3) directly accounts for the measurement origin uncertainty, i.e., which measurement comes from which speaker by using the probabilistic data association (PDA) technique with the IMM estimator. We demonstrate that a network of arrays combined with the estimation technique widely used in multisensor multitarget tracking area provides a consistent and coherent way to reduce the uncertainty and ambiguity of measurements. The effectiveness of our approach is illustrated by extensive simulation study on tracking a single moving speaker and two closely-spaced speakers with a crossing segment in their trajectories.

Local search heuristics for multi-index assignment problems with decomposable costs

The multi-index assignment problem (MIAP) with decomposable costs is a natural generalization of the well-known assignment problem. Applications of the MIAP arise, for instance, in the field of multi-target multi-sensor tracking. We describe an (exponentially sized) neighbourhood for a solution of the MIAP with decomposable costs, and show that one can find a best solution in this neighbourhood in polynomial time. Based on this neighbourhood, we propose a local search algorithm. We empirically test the performance of published constructive heuristics and the local search algorithm on random instances; a straightforward iterated local search algorithm is also tested. Finally, we compute lower bounds to our problem, which enable us to assess the quality of the solutions found.

FGAs-Based Data Association Algorithm for Multi-sensor Multi-target Tracking

A novel data association algorithm is developed basal on fuzzy genetic algorithms (FGAs). The static part of data association uses one FGA to determine both the lists of composite measurements and the solutions of m-best S-D assignment. In the dynamic part of data association, the results of the m-best S-D assignment are then used in turn, with a Kalman filter state estimator, in a multi-population FGA-based dynamic 2D assignment algorithm to estimate the states of the moving targets over time. Such an assignment-based data association algorithm is demonstrated on a simulated passive sensor track formation and maintenance problem. The simulation results show its feasibility in multi-sensor multi-target tracking. Moreover, algorithm development and real-time problems are briefly discussed.

Performance limits of track-to-track fusion versus centralized estimation: theory and application

Track-to-track fusion is an important part in distributed multisensor-multitarget tracking. The centralized and distributed tracking configurations were studied in (H.Chen et al., Proc. of SPIE Conf. on Signal and Data Processing of Small Targets, vol. 4048, 2000) using simulated air-to-air scenarios, and in (K.C. Chang, et al, IEEE Transact. on Aerospace and Electronic Systems, vol. 33, no. 4, pp. 1271-1276, 1997) with analytical results based on /spl alpha/-/spl beta/ filters. The current work generalizes the results in the latter to the cases with more than 2 sensors. As the number of sensors increases, the performance of the distributed tracker is shown to degrade compared with the centralized estimation even when the optimal track-to-track fusion is used. An approximate track-to-track fusion is presented and compared with the optimal track-to-track fusion with performance curves for various numbers of sensors. These performance curves can be used in designing a fusion system where certain trade-offs need to be considered. Finally, these results are compared with simulation results for a realistic air-to-air encounter scenario.

N-Tokyoites’ loop-line commuter problem

We present an O( n 2) order algorithm to an n-Tokyoites’ loop-line commuter problem. The n-Tokyoites’ loop-line commuter problem comprises a special class of the more general Gilmore–Gomory weighted bipartite matching problem where weights assigned to arcs are given in terms of integrals of some functions. The algorithm of O( n 2) complexity developed is faster than the more popularly used Hungarian-type O( n 3) algorithms (Naval Res. Logist. Quart. 2 (1955) 83; Management Sci. 12 (1964) 578) applicable to the more general weighted bipartite matching problem, but is slower than the original, more restricted Gilmore–Gomory O(n log n) algorithm (Oper. Res. 12 (1964) 655). The algorithm we have developed allows to impose some novel angular constraints which find an immediate application not only to the n-Tokyoites’ loop-line commuter problem itself, but also to the data association problem involved in the multisensor–multitarget tracking process (Design and Analysis of Modern Tracking Systems, Artech House, Norwood, MA, 1999) and to the specifically defined Gilmore–Gomory's original TSP problem.

Fast data association using multidimensional assignment with clustering

We present a fast data association technique based on clustering and multidimensional assignment algorithms for multisensor-multitarget tracking Assignment-based methods have been shown to be very effective for data association. Multidimensional assignment for data association is an NP-hard problem and various near-optimal modifications with (pseudo-)polynomial complexity have been proposed. In multidimensional assignment, candidate assignment tree building consumes about 95% of the time. We present the development of a fast data association algorithm, which partitions the problem into smaller sub-problems. A clustering approach, which attempts to group measurements before forming the candidate tree, is developed for various target-sensor configurations. Simulation results show significant computational savings over the standard multidimensional assignment approach without clustering.

A Set of Greedy Randomized Adaptive Local Search Procedure (GRASP) Implementations for the Multidimensional Assignment Problem

The focal problem for centralized multisensor multitarget tracking is the data association problem of partitioning the observations into tracks and false alarms so that an accurate estimate of the ...

Fuzzy logic data correlation approach in multisensor–multitarget tracking systems

In this paper, a fuzzy logic data correlation approach for multisensor–multitarget tracking is proposed. This fuzzy correlation approach is developed based on fuzzy clustering means algorithm. The proposed approach is applied to a two- and a four-dimensional multisensor–multitarget tracking system using Monte Carlo simulations. Fuzzy system performance evaluation is presented to demonstrate the efficiency of the new approach. The computational complexity of this approach is also analyzed and compared to that of conventional fuzzy logic data association methods. Considerable improvement in terms of computational complexity and performance is achieved.

Derivation and analytic evaluation of an equivalence relation clustering algorithm

Clustering algorithms have been recently used in multitarget multisensor tracking (MMT) problems in order to reduce the size of the data association problem. This paper derives an equivalence relation (ER) clustering algorithm used in a MMT problem and briefly compares it to other clustering schemes such as the nearest neighbor method. The main contribution of this work is the analytical evaluation of ER clustering performance, in the context of multitarget multisensor tracking, as a function of the distance between targets, measurement probability density function, and cluster parameter.

Precision large scale air traffic surveillance using IMM/assignment estimators

We present the development and implementation of a multisensor-multitarget tracking algorithm for large scale air traffic surveillance based on interacting multiple model (IMM) state estimation combined with a 2-dimensional assignment for data association. The algorithm can be used to track a large number of targets from measurements obtained with a large number of radars. The use of the algorithm is illustrated on measurements obtained from 5 FAA radars, which are asynchronous, heterogeneous, and geographically distributed over a large area. Both secondary radar data (beacon returns from cooperative targets) as well as primary radar data (skin returns from noncooperative targets) are used. The target IDs from the beacon returns are not used in the data association. The surveillance region includes about 800 targets that exhibit different types of motion. The performance of an IMM estimator with linear motion models is compared with that of the Kalman filter (KF). A number of performance measures that can be used on real data without knowledge of the ground truth are presented for this purpose. It is shown that the IMM estimator performs better than the KF. The advantage of fusing multisensor data is quantified. It is also shown that the computational requirements in the multisensor case are lower than in single sensor case, Finally, an IMM estimator with a nonlinear motion model (coordinated turn) is shown to further improve the performance during the maneuvering periods over the IMM with linear models.

A genetic algorithm based multi-dimensional data association algorithm for multi-sensor—multi-target tracking

The central problem in multitarget-multisensor tracking is the data association problem of partitioning the observations into tracks and false alarms so that an accurate estimate of true tracks can be found. The data association problem is formed as an N-dimensional ( N-D) assignment problem, which is a state-of-the-art method and is NP-hard for N ≥ 3 sensor scans. This paper proposes a new genetic algorithm for solving the above problem which is typically encountered in the application of target tracking. The data association capacities of the genetic algorithm have been studied in different environments, and the results are presented.

Fuzzy logic applications to multisensor-multitarget correlation

A consistent tactical picture requires data fusion technology to combine and propagate information received from diverse objects and usually vague situations. The information may be contained in two types of data; numerical data received from sensor measurements, and linguistic data obtained from human operators and domain experts. In real world situations, the numerical data may be noisy, inconsistent, and incomplete, and the linguistic information is imprecise and vague. To deal with these two types of data simultaneously, fuzzy sets and fuzzy logic provide a methodology to obtain an approximate but consistent tactical picture in a timely manner for very complex or ill-defined engineering problems. A functional paradigm for fuzzy data fusion is presented. It consists of four basic elements: (1) fuzzification of crisp elements, (2) fuzzy knowledge base derived from numerical input/output relations and humans, (3) fuzzy inference mechanism based on a class of fuzzy logic, (4) defuzzification of fuzzy outputs into crisp outputs for use by a plant. For real-time practical systems, the on-line determination of a fuzzy membership function from a given set of crisp inputs is vital. To this end, a methodology for estimating an optimal membership function from crisp input data has been implemented. This is based on the possibility/probability consistency principle as proposed by L.A. Zadeh. A relationship between the fuzzy membership function and the confidence level of statistical input data has been developed and it serves as a design parameter for fuzzification. This technique has been applied to a two-dimensional multisensor-multitarget tracking system. Fuzzy system performance evaluations have been presented. With simulated data in the laboratory environment, the simulation has been performed to evaluate the Mission Avionics Sensor Synergism (MASS) Systems. These results show better performance for the data correlation function using the fuzzy logic techniques.

Understanding GPS. Principles and applications

For multi-sensor multi-target tracking, traditional association-based methods treat data association and registration separately. However, they actually affect each other. The probability hypothesis density (PHD) filter has the distinct advantage that it avoids the complicated data association. In this paper, we propose an augmented state Gaussian mixture PHD (GM-PHD) filter with registration errors for multi-target tracking by Doppler radars. First, we construct the linear Gaussian dynamics and measurement model of the augmented state, which is comprised of target states and sensor biases. Then, related equations are derived when the standard GM-PHD filter is applied to the augmented state system. To effectively utilize Doppler measurements in the augmented state GM-PHD, the sequential processing method is adopted, i.e., updating target states and sensor biases with polar measurements first; and then updating sequentially target states with Doppler measurements; finally, computing weights with both polar and Doppler measurements. Simulation results show that the proposed filter is effective, and it has more robust performance in dense clutter.

Multitarget-Multisensor Tracking: Principles and Techniques [Book Review

Multitarget-Multisensor Tracking: Principles and Techniques [Book Review

Neural network optimization for multi-target multi-sensor passive tracking

In this paper, we review a number of neural network approaches to combinatorial optimization. We specifically address the difficult problem of localizing multiple targets using only passive sensors, i.e. the sensors detect only bearing angles. Thus, target positions must be found through triangulation. An efficient solution to this problem has been of particular interest in air defence applications. In this paper, we describe two different neural network based approaches for solving this passive tracking problem. In particular, we demonstrate the use of a Hopfield neural network to preface the subsequent development of the multiple elastic modules (MEM) model. The MEM model is presented as a significant extension to current self-organizing neural networks. We describe the unique features of the MEM model, including nonhomogeneous adaptive temperature field for escaping from poor local optima, and locking and expectation features used for dealing with dynamic real-world problems. Applications of the MEM model to other areas including computer vision, are also briefly described.

A Comparison of Parallel and Sequential Implementations of a Multisensor Multitarget Tracking Algorithm

Parallel and sequential implementations of the Multisensor Joint Probabilistic Data Association (MSJPDA) tracking algorithm are analyzed and compared. Two non-simulation techniques for comparison of multisensor probabilistic data association filters are developed and are used to compare tracking performance of the sequential and the parallel implementation of the algorithm. The non-simulation techniques are shown to accurately predict performance trends observed in simulations, that is the sequential implementation gives better tracking performance in terms of RMS position error and track lifetime. For the sequential implementation the question of processing order for different sensors is also briefly addressed. We also show that the sequential implementation is exponentially less computationally complex as the number of sensors increases. Thus, while sequential and parallel implementations are equivalent in multisensor filtering when no data association routine is needed, the sequential implementation gives superior tracking performance when data association is required.

Discussion of the Paper by Grenander and Miller

Discussion of the Paper by Grenander and Miller