基于计算机视觉的线性摄相机标定技术研究

通过建立基于一阶径向畸变的摄像机数学模型,提出了一种线性求解摄像机参数的标定方法,对摄像机模型中内部参数和外部参数进行逐步求解,且全部过程采用线性方法求解,从而改变了传统的摄像机标定依赖于非线性优化的缺点,避免了非线性优化的不稳定性.实验证明此方法具有较高的标定精度和较强的实用性.

Research on the Camera Linear Calibration Approach Based on the Computer Vision

In this paper a camera calibration approach is dealt with while radial distortion exists,a multi-step method is used which provides the exterior parameters and interior parameters solution,It can not only eliminate the nonlinear iterative calculation processes,but also replace the traditional nonlinear optimization and avoid obtaining unstable solution under common experiment conditions.It is proved that the new approach is very efficient and the accuracy is good enough.