摘要
摄像机定标技术在三维测量、计算机视觉、机器人等领域起着重大作用。分析了摄像机模型 ,提出了基于两阶段法和捆绑调整的摄像机定标方法 ,并通过模拟实验和真实实验进行了验证。标定块制作简单 ,标志点能够自动识别和匹配。标定中考虑了镜头畸变 ,在三维空间定标相对精度达1:10 0 0 0 。
Camera calibration plays an important role in 3 D measurements, computer vision, and robotics. This paper analyzes the camera model and proposes calibration methods based on the two stage method and bundle adjustment. Both simulated experiments and real data were used to verify the methods. The calibration pattern is easy to manufacture and the control points can be identified and matched automatically. Lens distortion is considered in the calibration, and the calibration accuracy is 1∶ 10 000 , satisfying the demand of high precision 3 D measurements.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2002年第4期481-483,493,共4页
Journal of Tsinghua University(Science and Technology)
