针对摄像机固定和共面约束条件下进行目标定位,提出了一种简单可行的全新算法。该算法避开了烦琐的摄像机标定这一过程,也不用精确地求解单应性矩阵。只是运用射影几何的性...

针对摄像机固定和共面约束条件下进行目标定位,提出了一种简单可行的全新算法。该算法避开了烦琐的摄像机标定这一过程,也不用精确地求解单应性矩阵。只是运用射影几何的性质,通过简单的几何关系,对非标定单摄像机拍摄的单幅图像,求解其从2D图像平面到真实的3D平面的非线性单映射关系。该算法为解决计算机视觉的测量、定位、三维重建等问题,提供一种全新的思路。 -against cameras fixed and coplanar binding conditions for target location, a simple new algorithm. The algorithm avoid the cumbersome process of the camera calibration, and it does not accurately be solved single matrix. Just use projective geometry of nature, by simple geometry, orders for non-standard video camera in a single image. Solving its two D images from two-dimensional to three real plane D nonlinear single map. The algorithm to solve the measurement of computer vision, positioning, three-dimensional reconstruction, with a brand new idea.

一种非标定摄像机的定位新算法.PDF