As target splitting is not considered in the initial development of δ-generalized labeled multi-Bernoulli (δ-GLMB) filter, the scenarios where the new targets appearing conditioned on the preexisting one are not readily addressed by this filter. In view of this, we model the group target as gamma Gaussian inverse Wishart (GGIW) distribution and derive a δ-GLMB filter based on the group splitting model, in which the target splitting event is investigated. Two simplifications of the approach are presented to improve the computing efficiency, where with splitting detection, we need not to predict the splitting events of all the GGIW components in every iteration. With component combination applied in adaptive birth, a redundant modeling for a newborn target or preexisting target could be avoided. Moreover, a method for labeling performance evaluation of the algorithm is provided. Simulations demonstrate the effectiveness of the proposed approach.
Splitting and combination are two important events of group target motion. However, the existing tracking approaches for group target splitting and combination events suffer the problems of high-computational cost and low accuracy. Under the random finite set framework, with target extent modeled by random matrix, the algorithms for group target splitting and combination tracking based on δ-generalized labeled multi-Bernoulli filter are researched. Three classical splitting modes of group target are discussed. With appropriate splitting criteria developed, e.g., the setting of the splitting gate, the chosen of the splitting dimension, the compensation of the subgroup's centroid position, and so on. According to the characteristics of each mode, the efficiency and the accuracy of the algorithm for group target splitting event are improved. The group combination approach is derived, where the representation of labels under the tack complicatedly changed condition, e.g., the group splitting and combination events jointly exist are given. With the velocity combination criterion established according to the target motion trend, a decreased sensitivity of the algorithm for target splitting event is avoided. The results show that the proposed algorithms have improved the tracking performance for group target splitting and combination events.INDEX TERMS δ-generalized labeled multi-Bernoulli, group target tracking, splitting, combination, gamma Gaussian inverse Wishart.
In order to improve the estimation performance of interacting multiple model tracking algorithm for group targets, the expected-mode augmentation variable-structure interacting multiple model (EMA-VSIMM) and the best model augmentation variable-structure interacting multiple model (BMA-VSIMM) tracking algorithms are presented in this paper. First, by using the EMA method, a more proper expected-mode set has been chosen from the basic model set of group targets, which can make the selected tracking models better match up to the true mode. The BMA algorithm uses a fixed parameter model of different structures to constitute a candidate model set and selects a minimum difference model from target state as the present extended model from the set of candidates at real time. Second, in the filtering process of VSIMM, the fusion estimation of extension state is implemented by the scalar coefficients weighting method, where weight coefficient is calculated by the trace of the corresponding covariance matrix of extension state. The performances of the proposed EMA-VSIMM and BMA-VSIMM algorithms are evaluated via simulation of a generic group targets maneuvering tracking problem.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.