摘要
提出一种基于单应矩阵的摄像机自动标定算法。讨论摄像机焦距为恒定和任意变化两种情况下求解摄像机内参数的计算方法:论证空间平面诱导单应矩阵的性质,利用该性质不但能求出摄像机外参数,还可得到空间平面法向量和单应矩阵方程的比例因子。该算法在求解过程中不需要非线性迭代,可以直接获得解析解,实验表明该算法具有很好的准确性、普遍性。
Proposes an algorithm for auto-calibration of a structured light system based on homography. Discusses how to determine＇the intrinstic parameters when the focal length is invariable or variable and the property of the homography based on a space plane, proves some conclusions involved. By using the property, we can＇t only obtain the extrinstic parameters but aslo the space plane normal vector and the scale factor. Since all computations involved are matrix operations in the linear algebra, and a closed form solution can be obtained.Experimental results of sythetic dates as well as real image dates show that the algorithm is correct and flexible and consequently.
关键词
结构光系统
自动标定
摄像机参数
三维重建
Structured Light System
Auto-Calibration
Camera Parameters
3D Reconstruction
