首页 | 本学科首页   官方微博 | 高级检索  
     

基于特征点跟踪的运动目标接触时间估计方法研究
引用本文:李俊,杨新,张桂林,施鹏飞. 基于特征点跟踪的运动目标接触时间估计方法研究[J]. 控制理论与应用, 2001, 18(Z1): 85-88
作者姓名:李俊  杨新  张桂林  施鹏飞
作者单位:上海交通大学图像处理与模式识别研究所
基金项目:自然科学基金(60072026)资助项目.
摘    要:解决估计运动目标和静止观测者之间的接触时间(time-to-contact)的问题.首先定义了广义接触时间的概念,并提出了基于特征点跟踪的估计匀速运动目标接触时间的理论依据和利用特征线段估计接触时间的解决思路.随后,提出了一个结合Kalman滤波器的估计匀速运动目标和静止观测者之间接触时间的特征点跟踪方案,并讨论了特征点的选择准则、运动分割的方法、以及所采用的特征点跟踪的方法.最后,针对标定TTC的运动目标序列图像进行接触时间的估计实验,实验的结果是令人满意的.

关 键 词:运动分析  接触时间估计  特征点跟踪  运动分割
文章编号:1000-8152(2001)-0085-04
修稿时间:2001-05-10

Feature-Point Tracking Approach to Time-to-Contact Estimation for Moving Objects
LI Jun and YANG Xin. Feature-Point Tracking Approach to Time-to-Contact Estimation for Moving Objects[J]. Control Theory & Applications, 2001, 18(Z1): 85-88
Authors:LI Jun and YANG Xin
Abstract:This paper deals with how to estimate the Time to Contact between moving objects and static observer. At first, the general concept of Time to Contact was defined. Then, a feature tracking system integrated with the Kalman filter was proposed to estimate the Time to Contact between moving objects and static observer. And some discusses were given, including the scheme of selection of feature points while tracking, extraction of moving objects, and feature point tracking scheme, etc. At last, the experiments to estimate TTC for the TTC labeled image sequences were done, and the results were promising. The research shows useful for robot and vehicle navigation.
Keywords:motion analysis  time to contact estimation  feature point tracking  motion segmentation  
本文献已被 CNKI 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