THRESHOLD-BASED RESAMPLING FOR HIGH-SPEED PARTICLE PHD FILTER
THRESHOLD-BASED RESAMPLING FOR HIGH-SPEED PARTICLE PHD FILTER
复制标题
基于阈值的高速粒子 PHD 滤波器重采样
DOI：
10.2528/pier12120406
复制
发表时间：
2013
影响因子：
6.7
通讯作者：
Yu, Zhengde
中科院分区：
文献类型：
--
作者：
Shi, Zhiguo;Zheng, Yunmei;Bian, Xiaomeng;Yu, Zhengde
近年来，粒子概率假设密度（PHD）滤波已成为密集杂波环境下多目标跟踪的一个研究热点。然而，由于粒子PHD滤波是一种蒙特卡罗方法，计算复杂度很高，因此对实时性要求很高。提高粒子PHD滤波器的真实的实时性能的主要困难之一在于，恢复是粒子PHD滤波器全并行实现的关键，而恢复通常是一个顺序的过程。为了克服这一困难，本文提出了一种新的基于阈值的粒子PHD滤波器恢复方案，其中粒子权重都设置在适当的阈值以下。该特定阈值是使用粒子PHD滤波器的区别特征来确定的：权重更新中所有粒子的权重总和等于当前迭代中的总目标数量。该方案允许在粒子PHD滤波器的硬件设计中使用全流水线结构。理论分析表明，在典型的多目标跟踪（MTT）场景下，采用该方法的粒子PHD滤波器的时间复杂度比传统的系统估计方法降低了33%左右，而仿真结果表明，该方法在保持估计精度的同时，仍能保持基本不变的性能。
In recent years, particle probability hypothesis density (PHD) flltering has become an active research topic for multiple targets tracking in dense clutter scenarios. However, it is highly required to improve the real-time performance of particle PHD flltering because it is a kind of Monte Carlo approach and the computational complexity is very high. One of major di-culties to improve the real- time performance of particle PHD flltering lies in that, resampling, which is usually a sequential process, is crucial to the fully-parallel implementation of particle PHD fllter. To overcome this di-culty, this paper presents a novel threshold-based resampling scheme for the particle PHD fllter, in which the particle weights are all set below a proper threshold. This speciflc threshold is determined using a distinguishing feature of the particle PHD fllters: The weight sum of all particles in weight update is equal to the total target number in the current iteration. This proposed resampling scheme allows the use of fully-pipelined architecture in the hardware design of particle PHD fllter. Theoretical analysis indicates that the particle PHD fllter employing the proposed resampling technique can reduce the time complexity by 33% around in a typical multi-target tracking (MTT) scenario compared with that employing the traditional systematic resampling technique, while simulation results show that it can maintain the almost same performance of estimation accuracy.
登录
查看更多内容
用于粒子滤波器的低功耗内存高效重采样架构
影响因子：
2.3
作者：
Chen, Ji-Ming;Shi, Zhi-Guo;Chen, Kang-Sheng;Hong, Shao-Hua
通讯作者：
Hong, Shao-Hua
影响因子：
--
作者：
Zhenkai Zhang;Jianjing Zhou
通讯作者：
Zhenkai Zhang;Jianjing Zhou
DOI：
10.1109/icassp.2003.1202435
发表时间：
2003-04
期刊：
2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03).
影响因子：
--
作者：
M. Bolic;P. Djurić;Sangjin Hong
通讯作者：
M. Bolic;P. Djurić;Sangjin Hong
DOI：
10.2528/pier09010302
发表时间：
2009
期刊：
--
影响因子：
--
作者：
Y. Li;East China;Y. Gu;Z. Shi;K. S. Chen
通讯作者：
Y. Li;East China;Y. Gu;Z. Shi;K. S. Chen
DOI：
10.2528/pierb12041701
发表时间：
2012
期刊：
--
影响因子：
--
作者：
J. T. Wang;H. Q. Wang;Y. Qin;Z. Zhuang
通讯作者：
J. T. Wang;H. Q. Wang;Y. Qin;Z. Zhuang