摘要： 针对道路交通事故现场图自动绘制中的摄像机标定问题，提出一种平面模板摄像机标定方法。在畸变较小的图像中心区域求取初值，采用基于内部映射牛顿法的子空间置信域法求解部分参数，引入畸变模型，通过直线特征约束求得剩余参数。实验结果表明，该方法能简化标定过程，减少运算量，提高计算速度。

关键词: 平面模板, 摄像机标定, 图像中心区域, 子空间置信域法, 畸变模型, 直线特征

Abstract: Aiming at the camera calibration in the drawing of the scene for the road traffic accident, a method for camera calibration based on planar template is proposed in this paper. Initial values are computed in the image central area where the distortion is small, partial calibration values are calculated through a subspace trust region method based on the interior-reflective Newton method, by introducing distortion model and employing linear feature constrain approach, remaining parameters can be calculated. Experimental results show that the proposed method simplifies calibration process, reduces computation and improves computing speed.

Key words: planar template, camera standardization, image central area, subspace trust region method, distortion model, linear feature

中图分类号: 

• TP301.6