文档介绍:第卷第期
45 2 吉林大学学报(工学版)
年月
2015 3 Journal of Jilin University(Engineering and Technology Edition)
基于多特征融合的粒子滤波自适应目标跟踪算法
杨欣1,2,3,刘加1,周鹏宇1,周大可1
南京航空航天大学自动化学院南京东南大学自动化学院南京光电控制技术重
(1. , 210016;2. , 210096;3.
点实验室河南洛阳
, 471000)
摘要为解决用单一特征无法保持在复杂环境下跟踪的鲁棒性以及粒子数量增多导致的算
:
法效率低下的问题,选择多个特征融合的策略来保证跟踪的持续稳定,并自适应地调整每个特
征的权值来适应环境的变化;为提高算法的实时性,采用自适应的粒子数量。实验结果表明:
本文算法有效地解决了目标旋转、目标遮挡以及背景混淆等诸多问题,具有较高的鲁棒性。
关键词:自动控制技术;目标跟踪;粒子滤波;多特征融合;自适应
中图分类号文献标志码文章编号
:TP29 :A :1671-5497(2015)02-0533-07
DOI:
Adaptive particle filter for object tracking based on fusing multiple features
YANG Xin1,2,3,LIU Jia1,ZHOU Peng-yu1,ZHOU Da-ke1
( of Automation Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing210016,China;
of Automation,Southeast University,Nanjing 210096,China; Laboratory of Photoelectric Control
Technology,Luogyang471000,China)
Abstract:In target tracking using a single feature,the tracking robustness can not be guaranteed in the
case plex environments,and the increase in particle number will lead to in low efficiency of the
e these problems,a multiple feature fusion strategy is chosen to ensure
sustained and stable this strategy,the weights of features will be adjusted adaptively
to the environmental changes,and the number of particles is adap