Abstract: To get continuous scanning of the structured-light vision system which is based on the projector and camera, it is needed to calculate the space position relationship between any light plane of the projector and image plane of the camera. Also the relative position between optical centre of camera and projector requires to be obtained. This paper settles the camera intrinsic parameters. Four corner points in the calibration board are used as characteristic points, also their extrinsic parameters are computed by using the intrinsic parameters of the camera. So coordinates of four characteristic points in the camera coordinate system can be found. The coordinates of the four characteristic points in the projector coordinate system can be computed by using the parameters of the projector itself. Thus, the relative position between optical center of camera and projector is obtained. That’s the process of calibration of structured-light based on camera and projector. The angular distance between corner points on calibration board is measured by structured-light vision system that has been calibrated above. The maximum relative error is 0.277%, which indicates that the calibration algorithm can be used in structure-light vision system based on projector and camera.

Key words: structured-light vision, calibration method, camera calibration, 3D recovery

摘要： 为实现基于投影仪和摄像机的结构光视觉系统连续扫描，需要计算投影仪投影的任意光平面与摄像机图像平面的空间位置关系，进而需要求取摄像机光心与投影仪光心之间的相对位置关系。求取摄像机的内参数，在标定板上选取四个角点作为特征点并利用摄像机内参数求取该四个特征点的外参数，从而知道四个特征点在摄像机坐标系中的坐标。利用投影仪自身参数求解特征点在投影仪坐标系中的坐标，从而计算出摄像机光心与投影仪光心之间的相对位置关系，实现结构光视觉标定。利用标定后的视觉系统，对标定板上的角点距离进行测量，最大相对误差为0.277%，表明该标定算法可以应用于基于投影仪和摄像机的结构光视觉系统。

关键词: 结构光视觉, 标定方法, 摄像机标定, 三维恢复