基于贝叶斯推理的目标跟踪  

Target Tracking Based on Bayesian Inference

作　　者：徐涛[1] 曾涛[1] 崔智社[1] 

机构地区：[1]北京理工大学电子工程系,北京100081

出　　处：《计算机仿真》2004年第5期74-77,共4页Computer Simulation

摘　　要：该文描述了基于贝叶斯推理的目标跟踪算法,可应用于非线性、非高斯系统中。介绍了贝叶斯跟踪基本概念。讨论了单目标跟踪算法及在不考虑数据关联、考虑数据关联两种情况下的多目标跟踪算法。给出了每种跟踪算法应满足的假设条件及其递归方程。最后,简要介绍了在工程中应用的具体算法。The paper introduces the target tracking algorithms based on Bayesian inference, which can be applied in the systems of nonlinearity and non-Gaussianity. We review the concepts of Bayesian tracking. Single target tracking algorithm and multiple target tracking algorithms in two situations, namely Multiple Target Tracking without association and with association, are respectively discussed. Also it presents the assumptions and the recursions of every kind of tracking algorithms. At last, some algorithms applied in projects are briefly described.

关 键 词：目标跟踪 贝叶斯推理 假设条件 递归方程 

分 类 号：TN959.9[电子电信—信号与信息处理]