Calibration

To calculate the coordinates of the points on the surface of the scanned object we need a previous step of calibration. This procedure allows us to determine to intrinsic and extrinsic parameters for the camera and the multimedia projector we are using as scanning hardware.

The intrinsic parameters are the focal length (on both directions x and y), the coordinates of the center of projection and the distortion factor. The extrinsic parametrs are the rotation matrix and translation vector (R and T, respectively). With these parameters we can find the relative position between the camera and the projector in 3D space.

To do this, we need a plane in space that can be used as reference for both the camera and the projector. This plane can be identified if we use a checker board as pattern (a common procedure among calibration methods) and examine it with a camera. Then we need to reproduce this same pattern with the projector.

We developed a script that first calibrates the camera, by taking a picture of the checker board pattern. Then, we remove this pattern, leaving only a white background on the plane and then we calibrate the projector, reprojecting a check borad image on the same plane and capturing another picture of it.

The check board pattern as seen by the camera

Projected check board pattern

Using our calibration software we identify all the corners of the black and white squares of the pattern, first for the camera image, then for the projector. Our system runs a least squares algorithm (as the Tsai calibration method) retrieving the intrinsic parameters for both. With this information, the software returns the camera and projector positions in space, expressed by their rotation and translation parameters. Calibration parameters are saved for the active triangulation procedure.

Last update December 12, 2003