基于粒子滤波的复杂背景下的红外运动目标跟踪  

Infrared Moving Target Tracking Based on Particle Filter in Complex Background

作　　者：王海梅[1] 洪敏 WANG Haimei;HONG Min(School of Automation,Nanjing University of Science and Technology,Nanjing 210094,China)

机构地区：[1]南京理工大学自动化学院,南京210094

出　　处：《火力与指挥控制》2018年第5期78-81,86,共5页Fire Control & Command Control

基　　金："十二五"军队预研基金资助项目

摘　　要：粒子滤波是一种处理非线性和非高斯动态系统状态估计的有效技术,针对目标被严重遮挡或有相似干扰等复杂背景情况下红外运动目标跟踪问题,提出了一种基于目标灰度与运动特征的粒子滤波算法。该算法将带有空间信息的灰度模型与带有灰度信息的运动模型进行融合,得到一个联合观测模型,并将其用于粒子滤波跟踪框架。与经典粒子滤波算法相比,文中算法效率略有降低,但跟踪的准确性和鲁棒性却大大增强。Particle filter is an effective technique to deal with the state estimation of nonlinear andnon Gauss dynamic systems. Aiming at the problem of infrared moving target tracking in complex background, such as serious occlusion or similar interference, a particle filter algorithm based on targetgray and motion feature is proposed. In this algorithm, the gray level model with spatial information andthe motion model with gray level information are fused to obtaina combined observation model, whichis used in the particle filter tracking framework. Compared with the classical particle filter algorithm,the algorithm efficiency is slightly lower, but the accuracy and robustness of tracking is greatly enhanced.

关 键 词：红外目标 目标跟踪 粒子滤波 联合观测模型 

分 类 号：TN971[电子电信—信号与信息处理]