Terry L. Ogle

Performance assessment and comparison of various tracklet methods for maneuvering targets

A tracklet is the estimate of a target state or track that is equivalent to an estimate based only a few measurements. Typically, tracklets are considered to reduce the communications costs between sensors and remote global or fusion trackers. The literature includes several methods for computing tracklets. Some of the methods compute tracklets from measurements, while others compute tracklets from the sensor-level tracks. Some of the methods ignore or omit process noise from the modeling, while others methods attempt to address the presence of process noise. The tracking of maneuvering targets requires the inclusion of process noise. When a tracklet that was developed for nonmaneuvering targets (i.e., no process noise) is used for tracking maneuvering targets, the errors of the tracklet will be somewhat cross-correlated with data from other sensors for the same target, and it is referred to as a quasi-tracklet. Due to some important practical considerations, the impact of maneuvering targets on the performance of tracklets has not been thoroughly addressed in the literature. An investigation that includes the critical practical considerations requires computer simulations with realistic target maneuvers and pertinent evaluation criteria (i.e., computation of errors). In this paper, some of the practical issues concerning the use of tracklets for tracking maneuvering targets are discussed, and the results from a simulation study of the impact of target maneuvers on tracking with tracklets are given. The study considered a fusion tracker receiving tracklets from multiple sensors at dispersed locations and targets maneuvering with either random accelerations or deterministic maneuvers. Tracklets from measurements and tracklets from tracks were studied. Since process noise was added to sensor and fusion trackers to account for target maneuvers, the tracklet methods studied are technically quasi-tracklets. A novel technique is used to compare the performance of tracklets for targets maneuvering randomly with that for targets performing deterministic maneuvers.

Multiple target tracking from images using the maximum likelihood HPMHT

In this paper, we address the problem of passive tracking of multiple targets with the help of images obtained from passive infrared (IR) platforms. Conventional approaches to this problem, which involve thresholding, measurement detection, data association and filtering, encounter problems due to target energy being spread across multiple cells of the IR imagery. A histogram based probabilistic multi-hypothesis tracking (H-PMHT) approach provides an automatic means of modeling targets that are spread in multiple cells in the imaging sensor(s) by relaxing the need for hard decisions on measurement detection and data association. Further, we generalize the conventional HPMHT by adding an extra layer of EM iteration that yields the maximum likelihood (ML) estimate of the number of targets. With the help of simulated focal plane array (FPA) images, we demonstrate the applicability of MLHPMHT for enumerating and tracking multiple targets.

Tracking separating targets with a monopulse radar: idealized resolution

The tracking of separating objects with a monopulse radar is a particularly challenging problem in that the presence of merged measurements result in significant delays in the initiation of tracks on the new objects. Furthermore, the unknown inflation in the variance and the bias of the angle-of-arrival estimates result in extra tracks and hinders track continuity. This paper presents an algorithm for tracking separating objects with a monopulse radar. The algorithm combines a Neyman-Pearson hypothesis test for detecting merged measurements, a method of parsing merged measurements into two angle-of-arrival estimates, and a track initiation procedure that takes advantage of the parsed measurements. The new algorithm was implemented in a sophisticated computer simulation environment to evaluate the performance improvements provided by the new algorithm. The output of the simulation environment is used to illustrate the challenges associated with tracking separating objects with a monopulse radar.

Advancement of an algorithm

The Probabilistic Multi-Hypothesis Tracker (PMHT) was developed in the early 1990s by Roy Streit and Tod Luginbuhl. Since that time many advances and improvements have been made to this elegant algorithm that is linearly efficient in processing as the number of targets, sensors, and clutter increases. This paper documents the many advances to the PMHT by several different contributors over the past two decades. The history continues and looks as promising as ever for this algorithm as we present the latest advancement—the Maximum Likelihood, Histogram Probabilistic Multi-Hypothesis Tracker (ML-HPMHT)—and the exciting results of this potential game-changer in tracking unresolved, dim targets in highly cluttered environments. This new algorithm, which we are calling the Quanta Tracking algorithm, detects and tracks with high accuracy targets that are unresolved in pixels or range bins.

Quanta tracking algorithm for multiple targets with velocity

The Quanta Tracking (QT) algorithm is a fairly new algorithm that is showing very promising results tracking unresolved, dim targets in highly cluttered environments. Traditional detection and tracking approaches use thresholding and signal processing to declare measurements that are then fed into the tracker. The QT algorithm does this all organically in an optimal manner, called “track-before-detect”. The algorithm requires no thresholding of the data such that all of the data is utilized. In the latest paper on this algorithm, we wrote about accounting for stationary targets and not tracking them. In this paper, we outlined two approaches for not tracking stationary targets. The first approach, implemented previously, removed any tracks that were below a certain velocity threshold. This allowed the algorithm to process the energy from these “stationary” targets, but then removed them from the set of tracks reported. Now we have implemented a second approach that prevents these “stationary” tracks from forming during the processing of the data. This approach has an advantage in that the algorithm will not spend any resources on these undesired stationary targets. In this paper, we show results from this latest approach as well as compare these results against the previous approach for not tracking stationary targets.

Design of IMM estimators with nearly constant velocity modes for tracking maneuvering targets

In numerous studies, the interacting multiple model (IMM) estimator has been shown to be significantly better than a conventional single model Kalman filter for tracking maneuvering targets. However, the design and criteria for which the IMM achieves maximum improvement over the conventional single model Kalman filter have received only little study. In this paper, a method for the design of IMM estimators with two nearly constant velocity (NCV) models is developed to maximize the reduction in average error and ensure the most consistent error covariance over NCV Kalman filters designed to minimize the maximum mean squared error (MMSE). Equations that determine the process noise standard deviation for each NCV model (nonmaneuver and maneuver models) in the IMM estimator are given as a function of the deterministic tracking index for targets with sustained maneuvers and sensors that have constant data rates. The design equations are verified via single sensor, single target Monte Carlo simulations by varying the deterministic tracking index from 0.01 to 100. The results of this paper characterize the performance bounds of the IMM estimator with two NCV models as compared to an NCV Kalman filter and provide a method for engineers to easily design IMM estimators with NCV models so that the benefits over the NCV Kalman filter are maximized.

Tracking maneuvering targets with multiple biased sensors

When tracking maneuvering targets with multiple sensors, the communication of measurements between the sensors or central-level tracker provides the best responsiveness to target maneuvers. However, when sensors are biased and the biases are not statistically insignificant, the maneuver detection and response in the central-level tracker is degraded. Due to the Markov modeling of the maneuver process in interacting multiple model (IMM) estimator, this degradation of the central-level tracking is acute. In this paper, the challenge of using an IMM estimator for central-level tracking of maneuvering targets with biased sensors is illustrated. In order to reduce the effects of sensor biases on the central-level tracks, the mode likelihoods are evaluated at the sensor and communicated to the central-level tracker along with the measurements. The benefits of using sensor-level mode likelihood in the central-level tracker is demonstrated via Monte Carlo simulations.

Air autonomous vehicle benchmark

The term benchmark originates from the chiseled horizontal marks that surveyors made, into which an angle-iron could be placed to bracket (“bench”) a leveling rod, thus ensuring that the leveling rod can be repositioned in exactly the same place in the future. A benchmark in computer terms is the result of running a computer program, or a set of programs, in order to assess the relative performance of an object by running a number of standard tests and trials against it. The Benchmark simulation environments began in the 1990s and continue today with many different variants including the Ballistic Missile Defense (BMD), Integrated Air/Missile Defense (IAMD), Chemical Biological Defense (CBD), Multiple-Input, Multiple-Output (MIMO) Radar, Electronic Countermeasures (ECM), and Electronic Attack (EA) Benchmarks. This latest variant of a benchmark, Air Autonomous Vehicle (AAV), for target tracking applications models autonomous air vehicles flying over littoral regions tracking surface vessels. This paper covers the capabilities of the AAV Benchmark, the initial scenario that was developed, and results showing the tradeoff between two centralized tracking algorithms.

PMHT for fused tracking

Fusing data together for target tracking is a complex problem. There are two key steps. First, the raw observations must be associated with existing tracks or used to form new tracks. Once the association has been done, then the tracks can be updated and filtered with the new data. The updating and filtering is usually the easier of the two parts and it is the association that can lead to most of the complexity in target tracking. When associating data (either measurements or tracks or both) with existing tracks, the separation between the tracks is critical to how difficult the association decisions will be. If the tracks are widely separated then the association decisions can be relatively easy. On the other hand, when the tracks are closely spaced the association decisions can be very difficult or nearly impossible. When the tracks or measurements are in three dimensions (such as with active sensors) the association can be accomplished in all three dimension thus making an easier distinction of targets that may be very close in two dimensions, but distant in the third dimension. However, when there are only two dimensions (as for passive sensors) observed by a sensor, targets that are widely separated may appear to be very close or even unresolved. In this paper, we will discuss the issues involved with applying the Probabilistic Multi-Hypothesis Tracking (PMHT) algorithm to fusing either measurements or tracks from passive sensors.