| 吴迪.基于改进粒子滤波和自适应融合的目标跟踪[J].光电子激光,2015,26(5):960~968 |
| 基于改进粒子滤波和自适应融合的目标跟踪 |
| Target tracking based on improved particle filter and adaptive fusion strategy |
| 投稿时间：2014-10-07 |
| DOI： |
| 中文关键词: 目标跟踪 粒子滤波(PF) 迭代无迹卡尔曼滤波(IUKF) 关联系数 自适应融合 |
| 英文关键词:object tracking particle filter (PF) iterated unsecnted Kalman filter (IUKF ) correlation coefficient adaptive fusion |
| 基金项目:国家科技支撑计划(1214ZGA008)、国家自然科学基金(61263031)、湖南省重点学科建设(081101)、湖南省教育厅项目(09C263)、重庆市教委自然科学基金(KJ1400628)和湖南工程学院博士科研启动基金资助项目 (湖南工程学院 电气信息学院,湖南 湘潭 411104) |
|
| 摘要点击次数: 1214 |
| 全文下载次数: 8 |
| 中文摘要: |
| 针对复杂环境下视频目标跟踪精确度低的问题,提 出了一种基于混合迭代无迹粒子滤波(HI-UPF)和关联系数自 适应融合的目标跟踪算法。首先采用统计线性回归的方法对无迹变换进行优化,提出了HI- UPF,不 仅提升了滤波精度,而且有效降低了算法的时间消耗；其次基于关联系数,采用一种自适应 融合方法,实现了加性 融合和乘性融合的自适应切换,并根据关联系数提出一种改进的自适应加性融合方法。仿真 实验表明,本文方法对 于复杂条件下的目标跟踪具有较高的精度和较强的鲁棒性。 |
| 英文摘要: |
| In order to solve the problem of low a ccuracy of video target tracking technique under complex environment,a new track ing method based on hybrid iterated unsecnted particle filter (HI-UPF) and corr elation coefficients adaptive fusion scheme is proposed in this paper.First,the new algo rithm optimizes unscented transformation using the statistical linear regression method,in which way,t he new method not only improves the filtering precision,but also effectively reduces the time consumption of the pro posed algorithm.Second,a new adaptive fusion technique is taken out by using the correlation coefficients to realize the adaptive change between the product rule and weighted sum rule.Experiment of results show that the propo sed method is more reliable and robust than traditional method. |
| 查看全文 下载PDF阅读器 |
| 关闭 |
|
|
|
|
|