Font Size: a A A

Research On Multi-sensor Multi-target Tracking Technology

Posted on:2006-02-16Degree:MasterType:Thesis
Country:ChinaCandidate:W W HuFull Text:PDF
GTID:2168360155968880Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Multi-sensor multi-target tracking and data association are the kernel content of multi-sensor multi-target data fusion, which includes two parts: target tracking and data association. This technology is widely applied in the field of military and civil, which are interested universally.Data fusion is a new foreland technology field, which combines many traditional subjects and burgeoning subjects. This thesis expatiates on the meaning, theory foundation, actual technology and research status of data fusion, which is a newly burgeoning subject dealing with information. This thesis also applies a survey in the data fusion research field and the focus of this thesis is multi-target tracking field of the data fusion.Considering the problem of the maneuvering.target tracking, firstly, this paper introduces basic problems about it, and discusses Interacting Multiple Model (IMM) algorithm in detail. It is showed theoretically that Fixed Structure Interacting Multiple Model (FSIMM) has limitations.The paper then presents Variable Structure Interacting Multiple Model (VSIMM) algorithm, which overcomes the limitations through using a variable set of models to select a better set of models. The new algorithm and the standard IMM are comprised by Monte- Carlo simulation. The results show that this algorithm not only improves the tracking accuracy for maneuvering targets, but also has moderate computation.On the problem of multi-sensor multi-target tracking in the clutter environment, this paper introduces several data association algorithms and their foreground firstly, and discusses a data association algorithm based on clustering. In this algorithm, it clusters track using FCM (Fuzzy C-Means) algorithm, and decides track quality according to the first step. The track that has highest quality is correlation. So the track is correlated using current and historical information,and correct correlation probability is advanced. The simulation result indicates that algorithm is efficient.The multi-target tracking theory on the condition of multi-sensor is discussed in this thesis and the future develop direction of multi-target tracking was forecasted.
Keywords/Search Tags:data fusion, multi-target tracking, data association, IMM, FCM
PDF Full Text Request
Related items