[1]罗翔,倪受东,颜景平.基于CAD模型的立体视觉目标运动估计[J].东南大学学报(自然科学版),2002,32(3):397-401.[doi:10.3969/j.issn.1001-0505.2002.03.019]
 Luo Xiang,Ni Shoudong,Yan Jingping.Kinetic parameters estimation of objects based on the CAD models by a stereo vision system[J].Journal of Southeast University (Natural Science Edition),2002,32(3):397-401.[doi:10.3969/j.issn.1001-0505.2002.03.019]
点击复制

基于CAD模型的立体视觉目标运动估计()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
32
期数:
2002年第3期
页码:
397-401
栏目:
自动化
出版日期:
2002-05-20

文章信息/Info

Title:
Kinetic parameters estimation of objects based on the CAD models by a stereo vision system
作者:
罗翔1 倪受东2 颜景平1
1 东南大学机械工程系,南京 210096; 2 南京工业大学机械工程学院,南京 210037
Author(s):
Luo Xiang1 Ni Shoudong2 Yan Jingping1
1 Department of Mechanical Engineering, Southeast University, Nanjing 210096, China
2 College of Mechanical, Nanjing University of Industry, Nanjing 210037, China
关键词:
机器人视觉 运动参数估计 扩展卡尔曼滤波
Keywords:
robot vision kinetic parameters estimation extended Kalman filters
分类号:
TP24
DOI:
10.3969/j.issn.1001-0505.2002.03.019
摘要:
提出了一种基于CAD模型的目标运动估计和目标轨迹实时生成方法.分为两步:首先采用立体视觉系统由图像特征点求出目标特征点在每个采样周期的的空间位置; 然后由其匹配模型的特征点集,通过扩展卡尔曼滤波方法估计目标运动参数.有了上述参数,就可以设计与伺服系统带宽匹配、消除图像处理延时影响的目标轨迹生成器.文中对目标的运动估计算法进行了仿真研究,结果表明该算法是可行的.文章创新之处在于推导了双目立体视觉系统的运动估计算法.
Abstract:
A scheme of kinetic parameters estimation and real-time track occurrence of objects based on the CAD models is investigated. The scheme has two steps. Firstly, the 3D positions of feature points of objects are obtained in every sample interval by a two-camera stereo vision system. Secondly, an extended Kalman filter is employed to estimate the kinetic parameters. With these parameters, a real-time track occurring rule is presented to match the bandwidth of the robot servo system and eliminate the influence of the delay involved in vision system. A simulation of kinetic parameters estimation of an object demonstrates the feasibility. The algorithm of kinetic parameters estimation by stereo vision systems is original.

参考文献/References:

[1] Hutchinson S,Hager G D,Corke P I.A tutorial on visual servo control[J].IEEE Trans Robot Automat,1996,12(5):651-670.
[2] Papanikolopoulos N P,Khosla P K,Kanade T.Visual tracking of a moving target by a camera mounted on a robot:a combination of control and vision[J].IEEE Trans Robot Automat,1993,9(1):14-35.
[3] 郑南宁.计算机视觉与模式识别[M].北京:国防工业出版社,1998.248-277.
  Zhen Nanning.Computer vision and pattern recognition[M].Beijing:Publishing House of Defence Industry,1998.248-277.(in Chinese)
[4] Janabi-Sharifi F,Wilson W J.Automatic selection of image features for visual servoing[J].IEEE Trans Robot Automat,1997,13(6):890-902.
[5] Wilson W J,Hulls C W,Bell G S.Relative end-effector control using Cartesian position based visual servoing[J].IEEE Trans Robot Automat,1996,12(5):684-696.
[6] Blostein S D,Huang T S.Error analysis in stereo determination of 3-D point position[J].IEEE Trans Pattern Anal Machine Intell,1987,PAMI-9(6):752-765.
[7] Arun K S,Huang T S,Blostein S D.Least-squares of two 3-D point sets[J].IEEE Trans on Pattern Anal Machine Intell,1987,PAMI-9(6):698-700.
[8] Goodwin G C,孙贵生.自适应滤波、预测与控制[M].北京:科学出版社,1992.200-246.
  Goodwin G C,Sun Guisheng.Adaptive filtering,forecast and control[M].Beijing:Publishing House of Science,1992.(in Chinese)

备注/Memo

备注/Memo:
基金项目: 211工程资助项目(2102002001).
作者简介: 罗翔(1968—),男,讲师,博士,luox@jlonline.com; 颜景平(联系人),男,教授,博士生导师.
更新日期/Last Update: 2002-05-20