Ship Targets Association Based On Multi-source Electronic Information

Targets association aims at determining whether the different information comes from the same object or not,which is the basis of multi-source information fusion.Under the background of using electronic reconnaissance satellite for maritime surveillance,this paper studies the problem of ship targets association based on multi-source electronic information.The main contributions are listed as followed.1.This paper presents the application value and theoretical significance of ship targets association based on multi-source electronic information.The current research on targets association is summarized,and the difficulties and the key issues to be resolved are analyzed.2.In the research of ship targets association based on radar electronic information with different measurement precision,this paper proposes a targets association algorithm based on attribute and fuzzy inference.Firstly,to pick out the ships,the data is clustered with attribute and processed with spatiotemporal constraints.Then,ships' radar emitter is recognized to construct the initial association measurement.Finally,based on different kinds of appropriate auxiliary information,fuzzy inference is used to improve the association accuracy.Simulation shows that this method can deal with ship association with strong false-alarm.3.In the research of ship targets association based on multi-satellite cooperating to finish maritime surveillance,association algorithm based on attribute and particle filter is proposed.Firstly,ship targets association algorithm based on attribute and fuzzy inference is used to pick out the tracking targets.Then,with the attribute,particle filter is applied to pre-estimate ship's location to associate the measurement.Finally,the result of algorithm based on attribute and particle filter can be certified.Simulation shows that this method can deal with the association problem on multi-satellite electronic information.4.In the research of ship targets association based on radar and communication electronic information,topological characteristics are applied to solve the problems,such as low location accuracy,high false-alarm interference and great undetected error rate.Consider the ship group as a point pattern,point pattern match algorithm can solve the association problem.Firstly,to be robust to noise and outliers,local topological characteristics are proposed.Then,based on local topological characteristics and probabilistic relaxation labeling,the correspondence is calculated to transform the point pattern.Finally,the thin plate spline-robust point matching is applied to obtain more correct correspondence.The experimental result shows the proposed method is robust to rigid transformation,noise and outliers and can solve the ship group association though the data is low location accuracy,high false-alarm interference and great undetected error rate.