A variational approach to problems in calibration of multiple cameras
A variational approach to problems in calibration of multiple cameras
复制标题
解决多相机校准问题的变分方法
DOI：
10.1109/cvpr.2004.24
复制
发表时间：
2004
期刊：
影响因子：
--
通讯作者：
A. Yezzi
中科院分区：
文献类型：
--
作者：
Gözde B. Ünal;A. Yezzi
本文讨论了用变分方法标定摄像机参数的问题。本文解决了广角/廉价相机镜头严重的镜头畸变问题。如果不考虑相机的畸变效应，会导致不准确的3D重建和几何测量。第二个问题是由相机响应的变化引起的颜色校准问题，这导致不同的颜色测量，并影响依赖于这些测量的算法。我们提出了基于变分思想的多视点立体技术来解决这些校准问题。为了降低这类算法的计算复杂度，我们利用过程中使用的校准对象的先验知识，进化出这样的三维模型对象的姿态、方向和比例参数。我们推导了畸变系数的演化方程和相机的颜色校准参数，并给出了实验结果，证明了它们的潜在用途。
This paper addresses the problem of calibrating camera parameters using variational methods. One problem addressed in this paper is the severe lens distortion in wide angle/inexpensive camera lenses. The camera distortion effects lead to inaccurate 3D reconstructions and geometrical measurements if not accounted for. A second problem is the color calibration problem caused by variations in camera responses which results in different color measurements and affects the algorithms that depend on these measurements. We present multi-view stereo techniques based on variational ideas to address these calibration problems. To reduce computational complexity of such algorithms, we utilize a prior knowledge on the calibration object which is used in the process, and evolve the pose, orientation, and scale parameters of such a 3D model object. We derive the evolution equations for the distortion coefficients, the color calibration parameters of the cameras, and present experimental results which demonstrate their potential use.
DOI：
10.1109/jra.1987.1087109
发表时间：
1987-08-01
期刊：
IEEE JOURNAL OF ROBOTICS AND AUTOMATION
影响因子：
--
作者：
TSAI, RY
通讯作者：
TSAI, RY