基于凸包的棋盘格角点自动识别与定位方法

 采用Harris方法提取角点，应用对称算子剔除棋盘格外圈角点与伪角点，基于灰度梯度方法精确提取棋盘格亚像素角点。在此基础上，提出了基于凸包的棋盘格角点分层识别与自动定位方法，实现了棋盘格角点物像坐标的自动匹配，从而可实现摄像机自动精确标定。实验结果表明，该方法具有较高的精度与可靠性，适合于摄像机在线自标定。

Automatic Extracting and Locating Checkerboard Corners Method Based On Convex Hull

By extracting checkerboard corners with Harris algorithm, a symmetry algorithm was proposed to identify the all inner-corners and eliminate outer corners and false corners of the checkerboard form primal corners muster. The method based on gray gradient was adopted for accurately extracting checkerboard corners. To distinguish the outer loop corners and automatic locate the corners form the corners muster of the checkerboard, a new method based on convex hull was presented, which can loop by loop recognize and sort all the corners. As a result, the image dots can be automatically matched with its corresponding physical dots, thus, parameters of camera were automatically and accurately calibrated. Experiments shows the method has high precision and preferable robust, and is especially adapted for the automatic online camera calibration.