Computer Engineering and Applications ›› 2012, Vol. 48 ›› Issue (2): 182-184.

• 图形、图像、模式识别 • Previous Articles     Next Articles

Dim moving point target tracking algorithm based on multi-source measurements information fusion

ZHOU Tongchi1, Askar Hamdulla1, WANG Rongshuan2, HUANG Dongmei3   

  1. 1.Institute of Information Science and Engineering, Xinjiang University, Urumqi 830046, China
    2.Yalong Communication, Tianjin 300384, China
    3.Urumqi Railway Administration, Urumqi 830011, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2012-01-11 Published:2012-01-11

多源测量信息融合的弱点动目标跟踪算法

周同驰1,艾斯卡尔·艾木都拉1,王荣栓2,黄冬梅3   

  1. 1.新疆大学 信息科学与工程学院,乌鲁木齐 830046
    2.天津市雅浪通讯,天津 300384
    3.乌鲁木齐铁路局电务段,乌鲁木齐 830011

Abstract: In order to improve anti-jamming of a dim moving point target tracking system under complex backgrounds, based on the new fusion tracking frame(add the local processor), tracking algorithm which is multi-source measurements information fuzzy logic fusion is proposed in different signal-to-noise conditions: local processor judges the original measurements by CFAR and adaptive detection threshold, transmits the measurements to fusion center; measurements which are fused by the fuzzy logic method are transformed into single virtual?image sequence’s measurements; adopts PDA-Kalman algorithm to track target. Compared to the single sequence, experimental analysis and results show that the fusion tracking algorithm has high accuracy and good stability, avoids losing the single sequence track.

Key words: multi-Infrared(IR) image sequences, dim small target, measurements fusion, probabilistic data as associated filter

摘要: 在复杂背景下,为提高微弱点动目标跟踪系统的抗干扰能力,从不同信噪比的多红外图像序列出发,基于新的融合跟踪结构(增加局部处理器),提出了一种模糊逻辑的多源信息测量融合跟踪算法:各局部处理器对原始测量进行恒虚警和自适应检测,将判决后的测量送入融合中心,融合中心对测量作模糊逻辑判决融合,转化为虚拟单序列测量,采用PDA卡尔曼滤波算法跟踪。实验结果分析表明,该融合跟踪算法与单序列相比,具有较高的跟踪精度、稳定性,避免了单图像序列跟丢。

关键词: 多红外图像序列, 弱小目标, 测量融合, 概率数据关联滤波