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

基于改进SUSAN算法的箭环目标跟踪与测量
引用本文:闫旻奇,唐慧君,张变莲.基于改进SUSAN算法的箭环目标跟踪与测量[J].光子学报,2007,36(10):1933-1938.
作者姓名:闫旻奇  唐慧君  张变莲
作者单位:1. 中国科学院西安光学精密机械研究所,西安,710119;中国科学院研究生院,北京,100039
2. 中国科学院研究生院,北京,100039
3. 西安文理学院,物理系,西安,710065
摘    要:针对发射场火箭初始段漂移量的测量,提出一种带灰度梯度方向的SUSAN算子,实现了基于该算子与变模板相关跟踪算法的目标跟踪测量方法.首先在SUSAN特征检测原则的基础上,将对角点的检测转化为对边缘的检测,同时记录梯度方向,在剔除了图像噪音点之后,由具有特定方向信息的边缘像素点精确定位角点的坐标.该方法克服了标准SUSAN算法准确度低的弊病,使得对目标的提取准确度可以达到亚像素级,并增强了抗噪性能,实验证明该算法提取准确性高,运算量小,易于实现.

关 键 词:漂移量测量  目标跟踪  SUSAN  角点检测  相关匹配
文章编号:1004-4213(2007)10-1933-6
收稿时间:2006-02-27
修稿时间:2006-02-25

A New Approach to Moving Target Tracking Based on Improved SUSAN Detector
YAN Min-qi,TANG Hui-jun,ZHANG Bian-lian.A New Approach to Moving Target Tracking Based on Improved SUSAN Detector[J].Acta Photonica Sinica,2007,36(10):1933-1938.
Authors:YAN Min-qi  TANG Hui-jun  ZHANG Bian-lian
Abstract:A new algorithm to measure the drift of rockets in the initial launching stage is presented and developed.It is designed to combine the improved SUSAN detector and the correlation.In this method,the corner detector is improved to detect the edge.Meanwhile,the direction of Gradient is recorded.After filter,the coordinates of corners are calculated accurately by using the edge pixels with special direction.The precision and the anti-noise performance is improved by this method.The results show that the precision reaches sub-pixel,with few mistake in detecting,and this algorithm is easy to realize for its less operations.
Keywords:Drift quantity measurement  Target tracking  SUSAN  Corner detector  Correlation
本文献已被 CNKI 维普 万方数据 等数据库收录!
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