首页 | 本学科首页   官方微博 | 高级检索  
     

基于队列流水线的弱点目标轨迹跟踪算法
引用本文:王江安,康圣,叶利民.基于队列流水线的弱点目标轨迹跟踪算法[J].光电工程,2007,34(9):1-4,9.
作者姓名:王江安  康圣  叶利民
作者单位:海军工程大学,电子工程学院,湖北,武汉,430033
摘    要:在宗思光等人提出的基于多量级多向梯度表决融合的双波段目标检测算法基础上,针对连续帧图像虚警点的线性累加性,以至于不易于目标跟踪及后期处理等不足,文章依据弱点目标在相邻帧间的运动不可能有大的跳跃,信号强度不会突变的特性,提出了可同时跟踪多个目标的算法,并进行了相应的实验,根据仿真已及实拍图片的实验,结果说明该方案是可行的,离实际应用要求又近了一步,具有较高的研究参考价值.

关 键 词:目标跟踪  弱点目标  目标识别  队列流水线
文章编号:1003-501X(2007)09-0001-04
收稿时间:2006/10/20
修稿时间:2006-10-202007-07-10

Tracking algorithm for dim-point target based on pipelining-queue
WANG Jiang-an,KANG Sheng,YE Li-min.Tracking algorithm for dim-point target based on pipelining-queue[J].Opto-Electronic Engineering,2007,34(9):1-4,9.
Authors:WANG Jiang-an  KANG Sheng  YE Li-min
Affiliation:Electronic Engineerin, Naval University of Engineering, Wuhan 430033, China
Abstract:On the basis of multi-degree and multi-orientation gradient voting fusion algorithm which was brought forward by Zong Si-guang, the number of false-alarm was reduced to less than 10 in one picture. In order to filter more false-alarm and offer more dynamic character for recognition, a tracking algorithm was presented based on queue-pipelining. Some experiments were done by using simulation and real images, and the result proves that this method has low false-alarm probability. This algorithm can be applied to remote pre-warning and accurate tracking according to different trust thresholds.
Keywords:target tracking  dim-point target  target recognition  queue-pipelining
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号