基于改进粒子滤波的空间红外小目标跟踪方法_图文.docVIP

第24卷 第11期 2010年11月 电子测量与仪器学报 Vol. 24 No. 11 · 1007 · JOURNAL OF ELECTRONIC MEASUREMENT AND INSTRUMENT DOI: 10.3724/SP.J.1187.2010.01007 基于改进粒子滤波的空间红外小目标跟踪方法 项学智1,2 彭 宇1 韩志英1,2 席志红2 (1. 哈尔滨工业大学电气工程及自动化学院, 哈尔滨 150001; 2. 哈尔滨工程大学信息与通信工程学院, 哈尔滨 150001 摘 要: 为有效跟踪空间红外小目标, 常采用基于粒子滤波的检测跟踪算法。粒子滤波中, 防止粒子退化的两个关键因素是选择合理的建议分布和重采样算法。在对重采样算法研究的基础上, 提出了以有效粒子数作为阈值进行模型切换, 分别利用残差重采样和系统重采样的优势, 将两种采样模型有机结合, 从而得到一种自适应的残差系统重采样(RSR算法。实验结果表明, 所提出的改进重采样算法在时间和跟踪精度上均优于经典重采样算法, 有效地提高了空间红外小目标跟踪的效率和稳定性。 关键词: 重采样；粒子滤波；目标跟踪 中图分类号: TP391.4 文献标识码: A 国家标准学科分类代码: 510.1040 Tracking method of space IR small target based on improved particle filter Xiang Xuezhi1,2 Peng Yu 1 Han Zhiying 1,2 Xi Zhihong 2 (1. School of Electrical Engineering and Automation, Harbin Institute of Technology, Harbin 150001, China; 2. College of Information and Communication, Harbin Engineering University, Harbin 150001, China Abstract: For tracking the infrared small target in the space effectively, particle filter method is often used. In particle filtering, the two key issues are the choosing of proper proposal distribution and resampling for avoiding particles degradation. An adaptive residual and system resampling (RSR algorithm is proposed in this paper, which regarded effective particle number as the threshold to switch the resampling model. This method uses the advantages of residual resampling and system resampling and the two sample models are combined effectively. Experiments show that the improved algorithm could track space infrared small target accurately in time. The total result is better than traditional algorithms. The proposed method can improve the efficiency and stability of tracking of the space infrared small target. Keywords: resampling; PF; target tracking 1 引 言 随着人类对空间探索能力的增强, 空间防御技术已成为各航天国家日益关注的课题, 并逐渐成为对空间红外小目标的检衡量国防实力的重要指标。 测及跟踪一直是空间防御中不可缺少的环节。迄今为止, 线性滤波领域中, 在最小均方误差准则下, 卡但空间小目标距尔曼滤波(KF是最优的滤波算法[2]。 离传感器较远, 探测光波严重衰减, 被观测的目标具有一定的机动性, 此时, KF难以达到满意效果。1969年, Bucy、Sunahara 等人提出了扩展卡尔曼滤波(EKF , 其核心思想是将非线性系统线性化处