[GEM-dev] pix_fiducialtrack help?
IOhannes m zmoelnig
zmoelnig at iem.at
Mon Feb 13 22:49:53 CET 2006
james tittle wrote:
> ...looking thru additions I haven't kept up with as much as I should,
> I'm trying to understand how this object works ;-) The data file
hmm
best thing (imo) is to read the papers that are mentioned in the help-patch.
> loads, but in the help file I don't get any output from the ID, x, y,
do you have a valid target at hand?
the fiducial-tracking does not work with general objects but only with a
special set of hierarchically topological targets (which are described
via the data file); on the reacTIVision homepage you should find a link
to a pdf that contains the complete set of ~127 fiducials that can be
tracked with the given data file.
(http://www.iua.upf.es/mtg/reacTable/pdfs/fiducials.pdf)
if you don't have a target and it doesn't detect one, than this means,
that the algorithm is stable :-)
you can also draw a target by hand (draw all lines really with a fat
black marker; "#" are just black blobs)
this should give you ID#22
+===============+
| _ _ _ _ _ |
| | | | | | | |
| |-|-|-|-|_| |
| |_|_|_|_| |
| _ _ |
| # # # |_|_| |
| |
| # # |
+===============+
> or angle...? Also, there's a second inlet that is undocumented in the
> object: is it just for another way to open a file, or something else?
the 2nd inlet is there for historic reasons (i think) and should be
eliminated.
> The code seems to mention a necessary vector...
where?
mfg.asdr
IOhannes
More information about the GEM-dev
mailing list