View Full Version : Number of control points for DLT

03-09-1998, 05:57 AM
Dear colleagues:

Tomislav Pribanic just posted a fine summary of replies to his recent
question on the DLT. Ton van den Bogert mentioned in his reply that
there is not much improvement in the DLT when more than 16 points are
used, referring to the Chen et al. paper (see below). This is not what
we have found at Arizona State University. I think the number is closer
to 50. Please refer to my paper in the Journal of Biomechanics in
October 1995 (volume 28, pp. 1219-1224.) There Scott McLean (I know
there are two Scott McLean's active in the biomechanics community
today--this is the one that is at Iowa State University) and I showed
the accuracy of predicting "non-control" points increased with
increasing numbers of control points up to 60 (the limit of our
available control points). We used 48 other points as "non-control"
points (points whose precise 3-D locations are known but which were not
used in the calibration). In the paper we discuss the limitations of
the Chen paper.

We recently purchased from Peak Performance Technologies a custum DLT
calibration object with approximately 50 points because their standard
one with 24 is not sufficient in my opinion. I'd like to know what
others think. Perhaps this will stimulate a discussion on biomch-l?



Richard N. Hinrichs, Ph.D.
Dept. of Exercise Science
Arizona State University

-- Ton van den Bogert replied to Tomislav Pribanic:

************************************************** **********************************

. .

. .
The MDLT has only 10 calibration parameters (not 11), since it assumes
that the
u and v axis in image space are perpendicular. This is true for film
and video.
With MDLT, you need at least 5 calibration points, but more is
for reliability. Reference:

Chen, L., Armstrong, C.W., Raftopoulos, D.D., (1994) An investigation on
accuracy of three-dimensional space reconstruction using the direct
transformation technique. J. Biomech. 27:493-500.
[Good paper on performance of the DLT using different sets of
points. It is shown that significant improvement can be gained by
increasing the number of calibration points to 16. More than that
does not help much. Some discussion of nonlinearity and how to
correct for that.]

-- Ton van den Bogert
Department of Biomedical Engineering
Cleveland Clinic Foundation
************************************************** *****************************