摄像机图像畸变纠正技术

A Technology of Image Distortion Rectification for Camera

本文通过对摄像机的畸变原因、简化的摄像机模型的讨论,提出了使用解二元高阶非线性方程组-牛顿叠代法对畸变进行纠正的方法,并给出了具体的算法描述和MATLAB中求解的算法流程图。最后,对该算法的逼近收敛速度进行了测试。测试结果表明,该算法只需通过3～4次逼近算法,收敛误差可小于10-5,能够快速地完成对摄像机畸变的纠正,完全可以满足对摄像机畸变进行纠正的要求。

The main contents of this paper are described as following： camera distortion and the simphfied camera model are discussed. To compensate the camera optical distortion we presented a way by solving dual - nonlinear high - order equations, that is, Newtonian compensation algorithm or iterative compensation algorithm, and we also provided the practice arithmetic flow chart in MATLAB. Finally, we take a test for convergence rate of this algorithm. The results shows that it only takes three or four convergence procedures to fulfill the fast distortion compensation for camera image, and the convergence error could be decreased below 10^-5. So it can meet the needs for fast distortion compensation for camera image perfectively.