基于灭点径向一致性约束的摄像机内参数求解

Approach for Determining Camera Intrinsic Parameters Based on Radial Consistency Constraint of Vanishing Point

基于透视投影成像的摄像机针孔模型,分析二维图像中的灭点和世界坐标系下无穷远点之间的几何关系,找出灭点之间的相应关系,并提取出灭点的坐标信息。在径向一致性约束的条件下,提出求解摄像机内参数的方法。利用C++平台的OpenCV库(开放性计算机视觉函数库)得到的实验结果和直接线性变换(DLT)实验结果进行了比较验证,表明该算法能准确、可靠地估计摄像机的内参数,并与实际情况吻合较好。

Based on the perspective projection model of pinhole camera, the geometric relationship between the vanishing point of two-dimensional images and point at infinity of the world coordinate system is analyzed to find the corresponding relations between the vanishing points, and obtain the coordinates of vanishing points information. Under the conditions of radial consistency constraint, the method is approached the camera intrinsic parameters is approached. The experimental results, which are obtained the OpenCV （Open Computer Vision） library of C＋＋ and Direct Linear Transform （DLE）, are compared to verify that the algorithm can accurately and reliably estimate the camera＇s intrinsic, and greatly agree with actual condition.