摘要
为了寻找一种简单、有效、精度高,并且使用方便的摄像机现场标定方法,提出将已知相位分布的平面二维正弦灰度调制条纹图作为标定平面靶,通过傅里叶条纹分析方法计算出两个正交相位分布,由正交相位分布与平面坐标的对应关系可以建立图像坐标与2D平面坐标的一一对应关系。将此二维正弦平面靶在摄像机成像空间中放置不同的位置,并完成相位测量,然后根据2D共面参照物摄像机标定方法即可完成摄像机标定。这种方法避开了复杂的标定特征点提取,大大增加了2D标定数据量,提高了标定精度。实验和计算机模拟表明,此方法简单、可靠和精度高,仅需测量3幅图像即可达到较高的标定精度。
A crossed fringe pattern plane was used as the model plane for camera calibration. The local world coordinates of the model planes can be decided by the previously known phase distributions on the planes. So the relationships between the 2D image coordinates and the 2D world coordinates will be set up, and the calibration procedure will be done by plane-based camera calibration technique after measuring the phase distributions of the reference plane at several different orientations. This avoids the complex procedure of extracting the control points from images, produces mass-volume calibration data, and leads to the high-accuracy calibration results even with the lest orientations in need. Both computer simulation and real data confirm the proposed technique is reliable, high accurate and high-speed.
出处
《四川大学学报(工程科学版)》
EI
CAS
CSCD
北大核心
2007年第6期149-153,共5页
Journal of Sichuan University (Engineering Science Edition)
基金
国家自然科学基金资助项目(60527001)
关键词
傅立叶轮廓术
摄像机标定
三维测量
2D共面参照物
平面二维正弦条纹图
Fourier transform profilometry
camera calibration
3D measurement
2D plane-based object
planar crossed fringe patterns