<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">On 01/08/2014 04:57 PM,
<a class="moz-txt-link-abbreviated" href="mailto:Martin.Azkarate@esa.int">Martin.Azkarate@esa.int</a> wrote:<br>
</div>
<blockquote
cite="mid:9429_1389196632_52CD7558_9429_6383_1_OFD28E1C65.D5E57D19-ONC1257C5A.00572134-C1257C5A.0057A4A6@esa.int"
type="cite"><font size="3">Hi there,<br>
<br>
Happy new year to everybody.</font>
<br>
<br>
<br>
<font size="3">We did some investigation and the error is not in
the
component not finding the subject. The error comes from the fact
that the
Nexus software does not calculate the coordinate system for the
subject/segment
in Live mode.</font>
<br>
<br>
<font size="3">If you look at the subject_segment_definition image
attached
you will see there a coordinate system clearly defined for it.
We are using
a wand with 5 markers, 3 of them are labeled and used to define
the segment
coordinate system and 2 markers are left unlabelled.</font>
<br>
<br>
<font size="3">In the live_stream_data_SDKTest image you will see
a snapshot
of the Nexus software running live (no capture process started)
and another
window showing the output of a console application that is
provided as
a testing program with the Vicon SDK v1.2. Note in this window
that the
Global Translation and the Global Rotation Quaternion vectors
for the segment
are all zero (which is an invalid quaternion value). Basically
Nexus does
not calculate (in live mode) Global values for the segment but
just positions
of markers. See position vectors for (3) labeled and (2)
unlabeled markers
are properly calculated.</font>
<br>
<br>
<font size="3">If you look at the output image of the test_vicon
ruby
scripts (attached script with slight modification to include
another reader
for the unlabeled markers port) you will see that it shows NaN
for the
translation values of whole subject rigidbodystate type (wrong
quaternion
screws the transformation matrix values) but it shows the same
values we
could see in the live_stream_data image for the unlabelled
Marker #0 (or
Maker #1) position values.</font>
<br>
<br>
<font size="3">So, we've been told by some colleagues at ESTEC
that in
fact <b>we need to use the Vicon Tracker software to have this
in live
mode (the Nexus is not enough)</b>, however I believe that
when this component
was developed in first instance it was done using the Vicon
Nexus, or that's
what I was told. So maybe we are missing something or we need to
prepare
the Nexus software to calculate/provide this data in live mode
too. <b>Anyone
using the Vicon component nowadays can tell us how does
configure/prepare
the Vicon Nexus and define the subject/segments?</b></font>
<br>
</blockquote>
<br>
Hi Martin,<br>
<br>
I remember there is a way in the Nexus software to set the real time
kinematics calculation. It seems you are missing this step.<br>
I don't know why but I think it is not set up by default. In order
to do that:<br>
<br>
1. Go to the "System" tab inside the "Resources" windows (left part
of the e.g.: <font size="3">live_stream_data_SDKTest image)<br>
2. Click in the "Local Vicon System [100Hz]"<br>
3. Below in "Properties" sub window, in "Core Processor" field<br>
4. In "Processing level:" set it to "kinematic fit" in the
deployed list.<br>
<br>
I hope this is the way to set it up. Otherwise, it should be in
the Nexus documentation. It might change among different Nexus
versions.<br>
<br>
Javier.<br>
<br>
</font><br>
</body>
</html>