引用本文: | 邱志强,唐力铁,于起峰.用神经网络变易有效焦距的摄像机标定法[J].国防科技大学学报,2002,24(5):16-19.[点击复制] |
QIU Zhiqiang,TANG Litie,YU Qifeng.A Camera Calibration Method of Varying Effective Focal Length by Neural Network[J].Journal of National University of Defense Technology,2002,24(5):16-19[点击复制] |
|
|
|
本文已被:浏览 6402次 下载 5633次 |
用神经网络变易有效焦距的摄像机标定法 |
邱志强1, 唐力铁2, 于起峰1 |
(1.国防科技大学 航天与材料工程学院,湖南 长沙 410073;2.国防科技大学 理学院,湖南 长沙 410073)
|
摘要: |
针对传统摄像机标定中由于像差等非线性因素的影响造成结果精度和稳定性不高且算法复杂的特点,提出了用4层前向神经网络模型变易摄像机横纵向有效焦距的方法。给出了摄像机标定结果和用标定后的摄像机进行三维立体测量的结果,并与传统方法作了对比。结果表明,基于神经网络的摄像机标定方法可以获得比传统方法更高的精度和稳定性。 |
关键词: 神经网络 摄像机标定 有效焦距 |
DOI: |
投稿日期:2002-04-30 |
基金项目: |
|
A Camera Calibration Method of Varying Effective Focal Length by Neural Network |
QIU Zhiqiang1, TANG Litie2, YU Qifeng1 |
(1.College of Aerospace and Materials Engineering, National Univ. of Defense Technology, Changsha 410073,China;2.College of Science, National Univ. of Defense Technology, Changsha 410073, China)
|
Abstract: |
The traditional calibration method is always imprecise, unstable and computed costly because of the nonlinear effect lead by lens distortion. A new camera calibration method of varying effective focal length by neural network model is presented. Both calibration result and 3D measurement result show that the new method achieves more precision and stable result than classical method does. |
Keywords: neural network cameral calibration equivalent focal length |
|
|
|
|
|