IEEE Transactions on Pattern Analysis and Machine Intelligence
Synopsis of recent progress on camera calibration for 3D machine vision
The robotics review 1
Decomposition of transformation matrices for robot vision
Pattern Recognition Letters
| Hi-index | 0.00 |
The problem of determining the intrinsic and extrinsic parameters of a mobile camera is addressed. An optimal solution is presented which consists of the following steps: first, the camera is calibrated in several working positions, and for each position the corresponding transformation matrix is computed using a method developed by Faugeras and Toscani^8; next, optimal intrinsic parameters are searched for all positions; finally, for each separate position, optimal extrinsic parameters are computed by minimizing a mean square error through a closed form solution. Experimental results show that such a technique yields a spectacular reduction of calibration errors and a considerable gain relative to other existing on-site calibration techniques.