引用本文: | 朱良家,董国华,周宗潭,等.本质流形上的自适应运动估计.[J].国防科技大学学报,2004,26(2):79-83.[点击复制] |
ZHU Liangjia,DONG Guohua,ZHOU Zongtan,et al.Adaptive Motion Estimation on the Essential Manifold[J].Journal of National University of Defense Technology,2004,26(2):79-83[点击复制] |
|
|
|
本文已被:浏览 6888次 下载 6331次 |
本质流形上的自适应运动估计 |
朱良家, 董国华, 周宗潭, 王正志 |
(国防科技大学 机电工程与自动化学院,湖南 长沙 410073)
|
摘要: |
分析了计算机视觉中一种新的运动估计模型,将运动估计作为一个参数定义在“本质流形”上的非线性系统参数辨识问题。针对基于离散处理和连续处理方法得到的“本质流形”,提出了一种新的、统一的自适应推广卡尔曼滤波算法,在“本质流形”的局部坐标系上进行运动估计。仿真结果验证了该算法的正确性。 |
关键词: 运动估计 本质流形 自适应卡尔曼滤波 |
DOI: |
投稿日期:2003-10-19 |
基金项目:国家自然科学基金重点资助项目(60234030) |
|
Adaptive Motion Estimation on the Essential Manifold |
ZHU Liangjia, DONG Guohua, ZHOU Zongtan, WANG Zhengzhi |
(College of Mechatronics Engineering and Automation, National Univ. of Defense Technology, Changsha 410073, China)
|
Abstract: |
A novel formulation of the visual motion estimation problem is analyzed in terms of identification of the nonlinear implicit system with parameters on the so called “essential manifold”; and a new and unified implicit adaptive extended kalman filter (IAEKF) is proposed to perform the estimation in the local coordinates, for both discrete and continuous cases. The correctness of the algorithm is illustrated by numerical simulations. |
Keywords: motion estimation essential manifold IAEKF |
|
|
|
|
|