Farid Amoozegar

Neural-network-based target tracking state-of-the-art survey

Target tracking research has been of interest to several different groups of researchers from different perspectives. An event of perhaps greatest importance in the history and development of target tracking research is the new trend in the architectural revolution in current algorithms and techniques that are used for target tracking, i.e., the advent of neural networks and their applications to nonlinear dynamical systems. It is established in the literature that the mathematical complexity of the state-of-the-art tracking algorithms has gone far beyond the computational power of conventional digital processors. Since the introduction of Kalman filtering, several powerful mathematical tools have been added to target tracking techniques, e.g., probabilistic data association, correlation and gating, evidential reasoning, etc. All these methods have one thing in common: they track targets rather differently from the way natural systems do. It is rather difficult to come up with a sound mathematical proof and verification of the concept for different parallel distributed architectures that seem appropriate for a general class of target tracking applications. However, the volume of contributions within the last decade in the application of various neural network architectures to different classes of target tracking scenarios can not simply be ignored. Therefore, the various neural-network-based tracking algorithms that have been introduced since 1986 are classified and addressed and their common views as well as their differences in results and in architectures are discussed. The role of mathematics in each of these algorithms and the extent that conventional methods are used in conjunction with the neural-network-based techniques are also addressed.

Efficient algorithm for computing data association probabilities for multitarget tracking

In this paper, an effective method is proposed to compute the data association probabilities in the joint probabilistic data association (JPDA) method. In this method, the a posteriori probability of the origin of each measurement required in JPDA, is computed directly, based on an approximate formula. The Computational and memory requirements of this method decrease greatly compared to the original JPDA. Simulation results show that despite great reduction in computational complexity, the algorithm approximates an exact JPDA more accurately than previously proposed simplifying methods such as SPDA.

Survey of fuzzy logic and neural network technology for multitarget tracking

Adaptive techniques for multi-target tracking have primarily been based on prior assumptions for the target and its background distribution. The statistical distribution theory, on the other hand, demands more complex mathematical modeling, which turns out to be computationally intensive as well. It is hard to deny the role of distribution theory and probabilistic approaches to the Multi-Target Tracking (MTT) particularly within the last two decades. However, despite the strength of statistical techniques and Bayesian approaches, the number of sensor samples for accurate modeling of current highly dynamical targets and their complex maneuvering capabilities require rather unrealistic assumptions about target dynamics. Practical target maneuvers with todays technology can be so short in duration that constant and uniform acceleration models for several samples may easily result in loss of tracks. This means the target can be undetected for many samples while making sharp turns. In recent years, there has been a paradigm shift toward fuzzy logic and neural network techniques. The membership functions of a fuzzy controller and nonlinear mapping capability of a trained neural network have made these two different technologies a viable combined system. The objective of this paper is to conduct a survey in the fuzzy logic technology as applied to target tracking and discuss its relation to neural networks when combined together.

Target tracking by neural network maneuver modeling

A new approach to tracking a maneuvering target using a neural network-based scheme is developed. The neural network models the target manoeuvre and assists a Kalman filter in updating its gains in order to generate correct estimates of target position and velocity. A performance evaluation of the target tracking scheme is conducted under various interesting scenarios. The parallel processing capabilities of trained neural nets are exploited in this application for realistically handling more input features to correct for the bias induced by the target manoeuvre.<<ETX>>

Multisensor multitarget tracking in 3D space using range and bearing measurements

In this paper the problem of estimation and tracking of geometric position of an object in a 3D space is considered using a network of sensors positioned at known points. Observations from each sensor include bearing and slant range of each object. Cramer-Rao error bound for estimation of target cartesian coordinates is derived and analyzed, in order to study the effects of measurement noise and sensor distribution in geometrical space. A basic structure for an adaptive algorithm is proposed for data acquisition and tracking.

Constant false alarm rate target detection in clutter: a neural processing algorithm

A serious degradation in detection probability of conventional Constant False Alarm Rate (CFAR) processors used in the automatic detection of radar targets results from a reduction in the number of available reference cells. Several factors such as any constraints on the radar system used (in terms of resolution and sampling time), presence of interfering targets and nonstationary clutter may contribute to the reduction in the number of reference cells. This paper presents a novel neural network-based CFAR detection scheme (referred to as NN- CFAR scheme) that offers robust performance in the face of loss of reference cells. This scheme employs a multilayer feedforward neural network trained by error backpropagation approach using the optimal detector as the teacher. The excellent pattern classification capabilities of trained neural networks are exploited in this application to effectively counter performance degradations due to reduced reference window sizes. In particular it is demonstrated that a neural network implementation of the CFAR detection scheme provides an efficient approach for accommodating more input parameters without increasing design complexity for countering the information loss due to reduced reference window size. Precise quantitative performance evaluation of the NN-CFAR scheme are conducted in a variety of situations that include both homogeneous and nonhomogeneous clutter backgrounds and the target detection performance is compared with that of the traditional CA-CFAR scheme to highlight the benefits.

Neural network detection and parameter estimation of airborne laser pulses through atmospheric transmission

While the laser radar systems have high performance at short ranges and low altitudes, the atmospheric effects have been the major constraints of detection and parameter estimation of laser pulses at long ranges and high altitudes. The turbulence which depends on different atmospheric states is hard to quantify due to the wavelength dependent effects of various conditions at different layers of the atmosphere. The turbulence may also be caused by interaction of the atmosphere with other objects, such as the vortex flow due to the aerodynamics of the air targets, or the nonlinear propagation characteristic of the high energy laser pulses. These adverse effects of the atmosphere have been limiting the usefulness of the laser radar systems for a wide range of applications. If the atmosphere is considered as a nonlinear media with nonuniform index of refraction, then it can be thought of as a nonlinear distributed lens under diffraction limited conditions. In this paper, a neural network modeling of the ionosphere layer is presented and the laser pulse is characterized by a set of input features. The transient CO2 laser pulses is simulated to transmit through the atmosphere to a satellite-borne receiver. The satellite receiver model is composed of three stages, i.e., the filtering and processing of the ionospheric propagated waveform, the envelope extraction and channel simulation, and the detection and parameter estimation. The received signal is then evaluated against the background noise through Monte Carlo simulations.

Data fusion and nonlinear tracking filter implementation using multilayer networks

One of the capabilities of multilayer neural nets that has not received much attention is the ability to efficiently fuse information of different forms for facilitating intelligent decision-making. In this paper the authors describe the capabilities and functionality of neural network algorithms for data fusion and implementation of nonlinear tracking filters. For a discussion of details and for serving as a vehicle for quantitative performance evaluations, the illustrative case of estimating the position and velocity of surveillance targets is considered. Efficient target tracking algorithms that can utilize data from a host of sensing modalities and are capable of reliably tracking even uncooperative targets executing fast and complex maneuvers are of interest in a number of applications. The primary motivation for employing neural networks in these applications comes from the efficiency with which more features extracted from different sensor measurements can be utilized as inputs for estimating target maneuvers. Such an approach results in an overall nonlinear tracking filter which has several advantages over the currently popular efforts at designing nonlinear estimation algorithms for tracking applications, the principal one being the reduction of mathematical and computational complexities.

Noise effect on the miss distance of a closed-loop neural-network-based radar homing missile for a tail-chase engagement

In recent years, parallel distributed processing has provided a new paradigm for algorithms, such as in missile guidance, which requires a high degree of computational efficiency as well as reliability and smaller size hardware. A problem of particular interest to the guidance literature is the closed-loop optical solutions that can be achieved on-board the missile. Furthermore, a desirable guidance scheme should be robust to low signal-to-noise conditions that generally arise in long-range applications. In this paper we shall present a neural network- based guidance scheme which provides a real-time optimal control on-board the missile with the inclusion of noise in the LOS angular rate data. The neural network is trained in an off-line session using optimal solutions obtained from an optimal control software resulting in a real- time closed-loop guidance method. The performance of the proposed scheme is then evaluated for different levels of SNR of the Line-Of-Sight (LOS) angular rate in a tail-chase engagement. In doing so, similar tests were conducted for the currently used closed-loop proportional navigation method and the potentially available technique of iterative optimal open-loop control with and without the presence of noise in the LOS angular rate. Although we did not include the noise in the missile/target dynamical model, the results indicate that the neural network-based scheme shows more robustness to low signal-to-noise situations as compared with traditional proportional navigation methods. This superiority is due, among other things, to the elimination of some of the restrictive, and in many cases unrealistic assumptions made in the derivation of most current guidance laws in use such as, for instance, unbounded control, simplified dynamics and/or aerodynamics, and non-maneuvering targets, to name a few.

Adaptive Kalman filter implementation by a neural network scheme for tracking maneuvering targets

Conventional target tracking algorithms based on linear estimation techniques perform quite efficiently when the target motion does not involve maneuvers. Target maneuvers involving short term accelerations, however, cause a bias (e.g. jump) in the measurement sequence, which unless compensated, results in divergence of the Kalman filter that provides estimates of target position and velocity, in turn leading to a loss of track. Accurate compensation for the bias requires processing more samples of the input signals which adds to the computational complexity. The waiting time for more samples can also result in a total loss of track since the target can begin a new maneuver and if the target begins a new maneuver before the first one is compensated for, the filter would never converge. Most of the proposed algorithms in the current literature hence have the disadvantage of losing the target in short term accelerations, i.e., when the duration of acceleration is comparable to the time period between the measurements. The time lag for maneuver modelings, which have been based on Bayesian probability calculations and linear estimation shall propose a neural network scheme for the modeling of target maneuvers. The primary motivation for employing compensation. The parallel processing capability of a properly trained neural network can permit fast processing of features to yield correct acceleration estimates and hence can take the burden off the primary Kalman filter which still provides the target position and velocity estimates.