中文摘要：

基于计算几何理论中Delaunay三角化原理,提出了一种对具有阵列网络结构的靶标特征点的图像坐标与世界坐标进行自动对应方法。通过对图像处理获取的特征点进行Delaunay三角化预处理,并去除冗余三角形,得到相邻特征点之间的几何连接关系,利用特征点之间的连接关系实现阵列网络结构的靶标特征点的图像坐标与世界坐标的对应。实验表明,提出的特征点图像坐标与世界坐标的自动对应方法实现简单,鲁棒性强,对镜头畸变不敏感,在拍摄倾角大、靶标被遮挡、内部特征点缺失的情况下对应结果仍很好,特别适合摄像机的在线自动标定。

英文摘要：

A simple correspondence approach in camera calibration is proposed based on Delaunay triangulation to automatically match the image coordinates with the world coordinates for feature points with regular grid network structure.The topological structure of the feature points was generated by Delaunay triangulation and pretreatment which removes the redundant triangles.The correspondence between the image coordinates and world coordinates of feature points was implemented by using the topological structure.The experimental results show that the proposed approach is simple,robust,and not sensitive to lens distortion.The approach can still effectively implement the feature point correspondence when the pattern is viewed from an oblique angle and under occlusion condition.It is most suitable for on-line camera calibration.