首页 | 本学科首页   官方微博 | 高级检索  
     

基于光流场技术的复杂背景下运动目标跟踪
引用本文:李宏,杨廷梧,任朴舟,李朝晖. 基于光流场技术的复杂背景下运动目标跟踪[J]. 光电工程, 2006, 33(10): 13-17
作者姓名:李宏  杨廷梧  任朴舟  李朝晖
作者单位:中国飞行试验研究院,陕西,西安,710089;中国飞行试验研究院,陕西,西安,710089;西安电子科技大学,陕西,西安,710072
摘    要:针对光电经纬仪在飞机起飞着陆段跟踪角速度大、背景复杂的情况,根据小波多分辨率技术,提出利用小波分层组成低频图像金字塔,利用光流技术进行匹配解算,实现由粗到精的特征点匹配跟踪。通过对飞行试验视频图像进行事后自动跟踪试验,结果表明,该算法能够稳定可靠跟踪多特征点目标,具有鲁棒性。

关 键 词:特征点跟踪  图像金字塔  小波多分辨率分析  光流场  目标跟踪
文章编号:1003-501X(2006)10-0013-05
收稿时间:2005-11-15
修稿时间:2006-04-24

Moving target tracking based on optical flow field analysis in complex background
LI Hong,YANG Ting-wu,REN Pu-zhou,LI Zhao-hui. Moving target tracking based on optical flow field analysis in complex background[J]. Opto-Electronic Engineering, 2006, 33(10): 13-17
Authors:LI Hong  YANG Ting-wu  REN Pu-zhou  LI Zhao-hui
Abstract:A novel automated method is developed to track aircraft takeoff and landing in a complex background. It involves three processes. Firstly, low frequency image pyramid is represented based on Wavelet multi-resolution and hierarchical theory. Then, an optical flow approach is used for matching of the pyramid representation. Finally the tracking method is completed through coarse-to-fine matching of feature point tracking. This method is evaluated by tracking flight video and found to track multiple targets with stability, reliability and robustness.
Keywords:Feature points tracking  Image pyramid representation  Wavelet-based multi-resolution rate analysis  Optical flow field  Target tracking
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