c++ - robust camera calibration -

i try , specific question, firstly doing project 3d image reconstruction using camera. have completed camera calibration not robust occlussions, such placing object on chessboard won't detect corners. there way of mopdifying program, or can use corners detected before placing object , how can use corners detect corners without occlussion.

i thinking of using corners representing 4 corners of chessboard, display , can place object on chessboard.

help appreciated in how can modify or use current camera calibration program deal occlusions i.e. placing object on chessboard.

for robust object detection, write program can detect larger number of features, , test whether sufficiently high percentage of features present.

for chessboard, try find 112 edges between squares, , see if line up. won't matter if miss few of 112.