基于双目视觉测量系统的投影仪标定方法

摘要

Calibration of the instrument parameters is the first problem in 3D vision measurement system, and the internal and external parameters of the projector calibration of structure light 3D measurement is an integral part of the system. For the present, the projector calibration still have problems, such as the precision is low, there is still a single method, it is poor operational. So we put forward a kind of projector calibration algorithm based on binocular vision, the projector is regarded as a reverse camera, the projector is used to capture images like a camera, establish the corresponding relation between the image plane and camera calibration, then obtain a correspondence between the camera and the projector image image with the help of Pole line principle. Then we can obtain the accurate relation between the projective surface and the projector image, we change the immature projector calibration into mature camera calibration. The tests show that the method can be achieved calibration requirements in accuracy.%在三维视觉测量系统中，对仪器参数的标定是首先需要解决的难题，而对投影仪的内外参数标定是三维结构光测量系统中不可或缺的一部分。就目前而言，投影仪的参数标定还存在着标定精度偏低、方法单一、可操作性差等问题。提出一种基于双目视觉的投影仪标定算法，将投影仪视为逆向相机，利用一个辅助相机捕获投影于不同位置标定平面，确立摄像机图像与标定平面之间的对应关系，然后利用极线原理得到摄像机图像与投影仪图像之间的对应关系，进而获得投影面与投影仪图像之间的准确关系，可将不成熟的投影仪标定转化为成熟的相机标定。实验结果表明，该方法在精度上能达到标定要求。