首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种新的红外成像末制导目标跟踪方法
引用本文:陈冰,赵亦工,李欣.一种新的红外成像末制导目标跟踪方法[J].光子学报,2014,38(11):3034-3039.
作者姓名:陈冰  赵亦工  李欣
作者单位:(西安电子科技大学 电子工程系模式识别与智能控制研究所,西安 710071)
摘    要:为了稳定跟踪导弹末制导阶段的红外目标,提出了一种基于尺度不变特征变换的红外目标跟踪算法.尺度不变性特征变换所提取的图像纹理特征具有尺度和旋转不变性,跟踪算法分别提取目标模板和待跟踪图像的尺度不变特征变换特征.根据最小欧氏距离准则提取目标模板与待跟踪图像间相匹配的尺度不变特征变换特征点对,利用该特征点对拟合反映两图像间映射关系的仿射模型,并据此估计目标中心位置及调整目标模板尺寸.仿真结果表明,跟踪算法能够较好地实现在导弹末制导阶段对红外地面杂波背景下目标的稳定跟踪,其跟踪准确度和稳定度优于传统方法.
关键词:末制导跟踪|尺度不变性特征变换|特征匹配|仿射模型

关 键 词:末制导跟踪  尺度不变性特征变换  特征匹配  仿射模型
收稿时间:2008-04-08

A New Approach to Target Tracking during Infrared Image Terminal Guidance
CHEN Bing,ZHAO Yi-gong,LI Xin.A New Approach to Target Tracking during Infrared Image Terminal Guidance[J].Acta Photonica Sinica,2014,38(11):3034-3039.
Authors:CHEN Bing  ZHAO Yi-gong  LI Xin
Institution:(Research Institute of Pattern Recognition and Intelligent Control,Xidian University,Xi’an 710071,China)
Abstract:In order to achieve robust tracking of an infrared target during terminal guidance,a new approach based on Scale Invariant Feature Transform (SIFT) is proposed.SIFT features are invariant to image scale and rotation,and express the image’s texture information.The SIFT features of template and query image are extracted respectively.And the matching keypoints between template and query image are extracted based on least Euclidean distance criterion.The matching keypoints are used to estimate target location and adjust the template size based on affine transformation model.Experimental results show that the proposed method provides more robust infrared target tracking in heavy background clutter during terminal guidance.
Keywords:Terminal guidance tracking  SIFT  Feature matching  Affine model
点击此处可从《光子学报》浏览原始摘要信息
点击此处可从《光子学报》下载免费的PDF全文
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