基于双目立体视觉的结构光测量技术

Structured Light Measurement Technique Based on Binocular Stereo Vision

目前的三维测量系统多数利用投影矩阵法求解出三维点坐标,但该方法未充分利用标定的参数,且矩阵之间计算复杂。为此,提出一种基于解析解的双目成像系统。该系统通过左右相机光线相交计算物体的三维坐标解析解,光线由相机的光心和成像平面上点的物理坐标计算得到。运用Steger算法提取条纹中心图像,并给出一种新的匹配点方法来匹配提取的条纹。实验结果表明,该系统通过48个标定点计算的误差为0.149 0 mm,从而验证了其有效性。

Most of the current three-dimensional measurement systems uses projection matrix method to solve the threedimensional point coordinates,butthis method does,not make full uses of the calibration parameters,and the caleultion between the matrix is complex. Therefore,a binocular imaging system based on analytic solution is proposed. The system calculates the analytic solutions of the object＇s three-dimensional coordinates by intersecting the left and right camera lights. The light is calculated from the camera＇s optical center and the physical coordinates of the points on the imaging plane. The Steger algorithm is used to extract the center of the stripe, and a new matching point method is proposed to match the extracted stripe. Experimental results show that the error of the system calculated by 48 calibration points is 0. 149 0 mm,which verifies the effectiveness of the proposed system.