<title>M-best SD assignment algorithm with application to multitarget tracking</title>

Abstract

In this paper we describe a novel data association algorithm, termed rn-best S D, that determines in O(mSkn3) time (m assignments, S lists of size ii, k relaxations) the rn-best solutions to an S D assignment problem. This algorithm is applied to the following problem. Given line of sight (i.e., incomplete position) measurements from S sensors, sets of complete position measurements are extracted, namely, the 1-st, 2-nd, . . . , rn-th best (in terms of likelihood) sets of composite measurements are determined solving a static S D assignment problem. Utilizing the joint likelihood functions used to determine the in-best S D assignment solutions, the composite measurements are then quantified with a probability of being correct using a JPDA-like technique. Lists of composite measurements from successive scans, along with their corresponding probabilities, are then used in turn with a state estimator in a dynamic 2 D assignment algorithm to estimate the states of the moving targets over time. The dynamic assignment cost coefficients are based on a likelihood function that incorporates the "true" composite measurement probabilities obtained from the (static) rn-best S D assignment solutions. We demonstrate this algorithm on a multitarget passive sensor track formation and maintenance problem, consisting of multiple time samples of line of sight (LOS) measurements originating from multiple (S = 7) synchronized high frequency direction finding sensors. Another significance of this work is that the rn-best S D assignment algorithm (in a sliding window mode) provides for an efficient implementation of a Multiple Hypothesis Tracking (MHT) algorithm by obviating the need for a brute force enumeration of an exponential number of joint hypotheses.

Beyond MAP Estimation With the Track-Oriented Multiple Hypothesis Tracker

“…As mentioned above, even with -scan pruning it is infeasible to sum over the entire space of hypotheses in (6). However, it is often the case in practice that the posterior probability mass is concentrated on a relatively small fraction of the hypotheses.…”

“…Variational message passing algorithms for graphical models offer an alternate approach to approximating the marginalization in (6). By taking advantage of conditional independencies present in the model, these algorithms can often obtain quick marginal estimates even when exact marginalization is very costly.…”

Multitarget multisensor closed-loop tracking

“…The basic types of detect to track algorithms discussed in the literature [1] are: The algorithms generally increase in complexity from SHT to PDA/JPDA to MHT [18]. Of course, there are many variations within each of these three general classes.…”