Sample source code for calibrating intrinsic parameter of a perspective camera
by Ryusuke Sagawa

This program is based on the following paper:

엧, ؍Nj
"Q̕s̊ϑɂJp[^̈ȃLu[V",
񏈗w, 񏈗w, No. 2007-CVIM-158, pp.??--??, 
, 3, 2007. 

To test an example of calibration, run test_paracalib.m on
MATLAB. The input data is in points.txt, which contains the
coordinates of feature point extracted from undistorted??.png. The
feature points are the corner of a building shown in the above paper.
The relative angle of the two points is 10.384 degrees in this
case. In the script, the main program, paracalib1.m, is called and the
intrinsic parameter K is computed. To visualize the error and
degeneracy, compute_surface.m is executed. The script requires
Symbolic Math Toolbox of MATLAB for visualization.

The format of point.txt is

    x1 y1 x2 y2

where the coordinates of two feature points are (x1, y1) and (x2, y2).

The undistorted images are generated from input??.bmp by the method
described in

Ryusuke Sagawa, Masaya Takatsuji, Tomio Echigo, Yasushi Yagi,
"Calibration of Lens Distortion by Structured-Light Scanning ", In
Proc. 2005 IEEE/RSJ International Conference on Intelligent Robots and
Systems, pp.1349--1354, Edmonton, Canada, August 2-5, 2005.

ҁ@, @, z@xv, ؁@Nj, "OCR[hp^[
𗘗pYc݂̕␳@", 摜̔FEV|WE
(MIRU2005), pp.174--180, 7, 2005.

