弱目标检测前跟踪技术研究综述  

Track-Before-Detect Technique for Dim Targets: An Overview

作　　者：杨威[1] 付耀文[2] 潘晓刚[1] 张志勇[1] 黎湘[2] 

机构地区：[1]国防科学技术大学九院,湖南长沙410073 [2]国防科学技术大学电子科学与工程学院,湖南长沙410073

出　　处：《电子学报》2014年第9期1786-1793,共8页Acta Electronica Sinica

基　　金：国家自然科学基金(No.61101181;No.401486)

摘　　要：该文阐述了弱目标检测前跟踪的基本原理,重点综述了近年来检测前跟踪技术的研究现状,包括动态规划、递归贝叶斯滤波、有限集统计学及直方图概率多假设跟踪等方法,对检测前跟踪技术的应用亦有相关介绍.最后在现有研究发展的基础上,着眼于提高针对弱目标的检测和跟踪性能及检测前跟踪算法鲁棒性的迫切需求,提出了检测前跟踪技术需重点关注和解决的若干问题,包括检测前跟踪算法的性能分析与评估、邻近弱目标检测前跟踪、机动弱目标的检测前跟踪、弱目标的多传感器融合检测和跟踪及弱目标的联合检测、跟踪与分类等方面.The basic principle behind the track-before-detect （TBD ）scheme was introduced ,and an overview of the studies on state-of-art TBD methods was then presented .The paper paid special attention to the following TBD methods ：dynamic program-ming ,recursive Bayesian filtering ,finite set statistics ,histogram probability multi-hypothesis tracking ,and also the relevant applica-tions .Finally ,some key issues to enhance the precision and robustness of dim target detection and tracking are indicated based on the progress of the existing research ,which deserve more attention of the researchers ’ for solution .These include：performance analysis and evaluation of the existing TBD methods ,the TBD solutions for closely-spaced and/or maneuvering dim targets ,multi-sensor fu-sion for the TBD problem ,joint target detection ,tracking and classification （JDTC ） for dim targets ,and so on .

关 键 词：弱目标检测 检测前跟踪 动态规划 粒子滤波 有限集统计学理论 

分 类 号：TN911[电子电信—通信与信息系统]