Tweet
Dear Colleagues:
In response to Andy Dainis' comment about his alternative to the DLT, I
would like to point out that I have advocated and used a method similar
to the one that Andy speaks of. It is known as the "NLT" method and was
developed by Dapena, Harman, & Miller in 1982. The equations are
clearly spelled out in two articles (J. Biomechanics, 15:11-19, 1982;
and J. Biomechanics, 18:163, 1985). In fact, the article I mentioned in
my posting of a few days ago (J. Biomechanics, 28: 1219-1223, 1995)
compared the NLT to the DLT and discusses the pros and cons of each
method. The beauty of the DLT over the NLT is the fact that you can get
quite accurate results without having to bother to do a separate
calibration of each camera for internal camera parameters (a weakness of
both the NLT and Andy's method). The advantages of these latter methods
is that they can be used with VERY LARGE control volumes and use much
simpler control objects (a single pole with 2 points on it will
suffice).
However, the bottom line is that the vast majority of researchers who do
3-D motion analyses today use the DLT (not these other methods), and
hence I was hoping to stimulate discussion on how to best use the DLT,
not the NLT or Andy's (presumably?) proprietary method. If, on the
other hand, the readership wants to expand to other 3-D methods, I'm
game for that too. I'd be glad to share my experiences with
extrapolating the DLT, for example. Thanks for everyone's input so
far. I'll keep a running tally for summary purposes later.
Regards,
--Rick
Richard N. Hinrichs, Ph.D.
Dept. of Exercise Science
Arizona State University
Hinrichs@asu.edu
In response to Andy Dainis' comment about his alternative to the DLT, I
would like to point out that I have advocated and used a method similar
to the one that Andy speaks of. It is known as the "NLT" method and was
developed by Dapena, Harman, & Miller in 1982. The equations are
clearly spelled out in two articles (J. Biomechanics, 15:11-19, 1982;
and J. Biomechanics, 18:163, 1985). In fact, the article I mentioned in
my posting of a few days ago (J. Biomechanics, 28: 1219-1223, 1995)
compared the NLT to the DLT and discusses the pros and cons of each
method. The beauty of the DLT over the NLT is the fact that you can get
quite accurate results without having to bother to do a separate
calibration of each camera for internal camera parameters (a weakness of
both the NLT and Andy's method). The advantages of these latter methods
is that they can be used with VERY LARGE control volumes and use much
simpler control objects (a single pole with 2 points on it will
suffice).
However, the bottom line is that the vast majority of researchers who do
3-D motion analyses today use the DLT (not these other methods), and
hence I was hoping to stimulate discussion on how to best use the DLT,
not the NLT or Andy's (presumably?) proprietary method. If, on the
other hand, the readership wants to expand to other 3-D methods, I'm
game for that too. I'd be glad to share my experiences with
extrapolating the DLT, for example. Thanks for everyone's input so
far. I'll keep a running tally for summary purposes later.
Regards,
--Rick
Richard N. Hinrichs, Ph.D.
Dept. of Exercise Science
Arizona State University
Hinrichs@asu.edu