首先自制一张标定图片，用A4纸打印出来，设定距离，再设定标定棋盘的格子数目，如8×6，然后利用cvFindChessboardCorners找到棋盘在摄像头中的2D位置，这里cvFindChessboardCorners不太稳定，有时不能工作，也许需要图像增强处理。



计算实际的距离，应该是3D的距离。我设定为21.6毫米，既在A4纸上为两厘米。



再用cvCalibrateCamera2计算内参，



最后用cvUndistort2纠正图像的变形。

-First, a self-calibration picture, with A4 paper, print it out, set the distance, then set the calibration number of the board of the lattice, such as 8 × 6, and then use cvFindChessboardCorners find the board in the position of the camera in 2D, where cvFindChessboardCorners less stable, sometimes does not work, may need to image enhancement. Calculate the actual distance, the distance should be 3D. I set 21.6 mm, both in the A4 paper to two centimeters. Then cvCalibrateCamera2 calculated internal reference, and finally with cvUndistort2 correct the distortion.