From H.Miri at 2007.hull.ac.uk Mon Jun 2 08:07:37 2008
From: H.Miri at 2007.hull.ac.uk (Hossein Miri)
Date: Mon, 2 Jun 2008 13:07:37 +0100
Subject: [Aria-users] gcc 3.4.2
Message-ID: <445800D8FD5D3D43878FD7D2842A3047898F62@EXCL2VS2.adir.hull.ac.uk>
Hi Matt,
In one of your documents, you have printed these instructions:
To install gcc-3.4.1, download its sources archive from:
http://gcc.gnu.org
Become a superuser, and untar the archive into:
/usr/src or other directory, just *not* /usr/local:
# cp gcc-3.4.1.tar.bz2 /usr/src/
# cd /usr/src
# tar -zxvf gcc-3.4.1.tar.gz
Which one should be downloaded: .bz2 or .gz?
I'm trying to install gcc 3.4.2 on Fedora 4 by following the same instructions
because I get the error message: g++ command not found (when I run MAKE in the ARIA examples).
But I'm confused as to where to download, which one, and how to install it.
Thanks,
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080602/6e69496e/attachment.html
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: not available
Url: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080602/6e69496e/attachment.ksh
From reed at mobilerobots.com Mon Jun 2 09:05:39 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Mon, 02 Jun 2008 09:05:39 -0400
Subject: [Aria-users] gcc 3.4.2
In-Reply-To:
References:
Message-ID: <4843F023.6020200@mobilerobots.com>
Hossein MIRI wrote:
> Which one should be downloaded: .bz2 or .gz?
Either one. bzip2 is smaller (more compressed).
To unzip .bz2 use:
tar xjf gcc-3.4.1.tar.bz2
To unzip .gz use:
tar xzf gcc-3.4.1.tar.gz
Reed
From deleon at atp.ruhr-uni-bochum.de Mon Jun 2 11:13:20 2008
From: deleon at atp.ruhr-uni-bochum.de (Plinio de Leon Canton)
Date: Mon, 02 Jun 2008 17:13:20 +0200
Subject: [Aria-users] How to measure the supplied motor current
Message-ID: <48440E10.3050403@atp.ruhr-uni-bochum.de>
Dear all,
I'm working with a Pioneer p3dx and I need to get the Current (or
Voltage) supplied to the
motors. Did anyone tried this before or have any idea how to do this?
Regards,
Plinio
From zeb at mobilerobots.com Mon Jun 2 11:16:04 2008
From: zeb at mobilerobots.com (Zeb Dahl)
Date: Mon, 02 Jun 2008 11:16:04 -0400
Subject: [Aria-users] How to measure the supplied motor current
In-Reply-To: <48440E10.3050403@atp.ruhr-uni-bochum.de>
References: <48440E10.3050403@atp.ruhr-uni-bochum.de>
Message-ID: <1212419764.1911.16.camel@localhost.localdomain>
Hello,
There is no direct way to measure motor currents built into the robot.
In order to measure current, you would need to add some sort of current
sensor to the motor leads. System voltage is reported in the standard
SIPs regularly, but that probably wouldn't help you too much.
Best,
-Zeb Dahl
On Mon, 2008-06-02 at 17:13 +0200, Plinio de Leon Canton wrote:
> Dear all,
>
> I'm working with a Pioneer p3dx and I need to get the Current (or
> Voltage) supplied to the
> motors. Did anyone tried this before or have any idea how to do this?
>
> Regards,
>
> Plinio
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail to:
>
> aria-users-leave at lists.mobilerobots.com
>
From hmiri at mac.com Mon Jun 2 11:44:06 2008
From: hmiri at mac.com (H . Miri)
Date: Mon, 02 Jun 2008 08:44:06 -0700
Subject: [Aria-users] g++
Message-ID: <41EAA149-011A-1000-836F-197EB8BF3EE8-Webmail-10018@mac.com>
Hi all,
I have two Linux boxes: a Ubuntu and a Fedora 4
On both, I always got the error message: g++ command not found
On my Ubuntu machine, I resolved this by using these two commands:
sudo apt -get install g++
sudo apt -get install build-essential
Now I can go into the ARIA Examples directory and MAKE the examples.
Does anyone know of any equivalent commands to use in Fedora 4 so that it installs g++ automatically? (without having to go through the troubles of downloading packages, extracting, untarring, etc.)
Cheers,
From eypcinar at gmail.com Tue Jun 3 05:00:08 2008
From: eypcinar at gmail.com (=?ISO-8859-1?Q?Ey=FCp_=C7INAR?=)
Date: Tue, 3 Jun 2008 12:00:08 +0300
Subject: [Aria-users] ArLocalizationTask functions
Message-ID: <2f25bbee0806030200n74c10682ve0038c1771211bdd@mail.gmail.com>
Hello all,
Does anybody of you may give some idea about the difference between
Arnl ArLocalizationTask::localizeRobotAtHomeBlocking () and
ArLocalizationTask::localizeRobotAtHomeNonBlocking ()
for NonBlocking it is said as explanation the task later localize the robot
at a map home position.. As far as I understood from the explanation of the
first function, it needs to collect some sensor data in order to achieve
initialization therefore I interpret that I must call it at some time after
the robot moves. However, this is not suitable for my program because I have
some usertasks and my program runs in synch. mode with robot.run()
When I use the second one, localization works but sometimes I get this
warning at the begining
Reseting all localization flags
and when i get this it seems the robot does not localize itself . Do you
know what "Reseting all localization flags" mean ?
Thanks for any idea
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080603/bd0d4a45/attachment.html
From H.Miri at 2007.hull.ac.uk Tue Jun 3 08:56:52 2008
From: H.Miri at 2007.hull.ac.uk (Hossein Miri)
Date: Tue, 3 Jun 2008 13:56:52 +0100
Subject: [Aria-users] (no subject)
Message-ID: <445800D8FD5D3D43878FD7D2842A3047898F6C@EXCL2VS2.adir.hull.ac.uk>
Hi everybody,
When I run wander.cpp this is what happens:
$ ./wander
./wander: error while loading shared libraries: /usr/local/Aria/lib/libAria.so: cannot restore segment prot after reloc: Permission denied
So, I checked to see if I could figure out the issue with this lib file:
$ cd /usr/local/Aria/lib
$ ls -l
total 24680
-rwxr-xr-x 1 root users 3166469 Oct 17 2005 libAriaJava.so
-rwxr-xr-x 1 root users 16699697 Jun 3 12:08 libAria.so
-rwxr-xr-x 1 root users 5347416 Oct 14 2005 libArNetworking.so
I did a bit of naughtiness and compared this to a working machine. This is what I (should) get:
total 24680
-rwxr-xr-x 1 root root 3166469 Oct 17 2005 libAriaJava.so
-rwxr-xr-x 1 root root 16699697 Jun 3 12:08 libAria.so
-rwxr-xr-x 1 root root 5347416 Oct 14 2005 libArNetworking.so
The fourth column is all root. Could this have anything to do with the error message I get? Any solutions?
Cheers,
---
Hossein MIRI (Joshua) - Ph.D. Student
Neural, Emergent, and Agent Technologies Research Group (NEAT)
Department of Computer Science, University of Hull, Kingston-upon-Hull
East Riding of Yorkshire, HU6 7RX, United Kingdom
http://www.hmiri.com
http://hjoshua.squarespace.com
---
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080603/e402c338/attachment-0001.html
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: not available
Url: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080603/e402c338/attachment-0001.ksh
From kennedy at MobileRobots.com Mon Jun 9 08:43:00 2008
From: kennedy at MobileRobots.com (William Kennedy)
Date: Mon, 09 Jun 2008 08:43:00 -0400
Subject: [Aria-users] problems with ARNL
In-Reply-To: <484C06F7.7040309@cs.cmu.edu>
References: <484C06F7.7040309@cs.cmu.edu>
Message-ID: <6.2.5.6.2.20080609084213.03691278@MobileRobots.com>
Hi, Paul:
We probably can work something out for you, but in the meantime,
let's let Aria-users see if they can help. Reed?
Bill
At Sunday 6/8/2008, Paul E. Rybski wrote:
>Hey Bill,
> I'm running into problems with our use of ARNL. We're running Ubuntu
>7.10 on our machines and we're getting very strange crashing problems
>that appear to be coming deep within ARIA and ARNL. I've tried building
>ARIA from source rather than using the Debian .deb package and that has
>fixed the crashing problems we were having with our AIRA-only software.
> However, since ARNL is distributed as binary-only I'm concerned that
>our only hope of getting it working might be to recompile it from
>source. So... my question to you is:
>
>Can we work out some sort of NDA (or use the existing ones I've got with
>you...) to get the ARNL source so I can compile it on my own machine and
>then distribute the binaries to the rest of our machines?
>
>Thanks,
>
>-Paul
From H.Miri at 2007.hull.ac.uk Mon Jun 9 09:54:18 2008
From: H.Miri at 2007.hull.ac.uk (Hossein Miri)
Date: Mon, 9 Jun 2008 14:54:18 +0100
Subject: [Aria-users] /usr/bin/ld: cannot find -lAria
Message-ID: <445800D8FD5D3D43878FD7D2842A3047898F7A@EXCL2VS2.adir.hull.ac.uk>
Hi Reed,
When I run Make in my Aria directory (not the one that comes with the package - the one that has my Prolog and C++ programs) I get the bunch below. The first part is obviously my files, but what is the error message at the end?
plld -c++ -I include -I/usr/local/Aria/include -I/usr/lib/pl-5.6.4/include/ -ldl -lpthread -lAria MyActionSonarConstantVel.o MyActionSonarAvoidSide.o MyActionSonarAvoidFrontNear.o MyActionSonarAvoidFrontFar.o MyActionSonarAvoidRear.o MyActionVisionConstantVel.o MyActionVisionAvoidFrontNear.o MyActionVisionAvoidFrontFar.o MyTrackObject.o MyHitObject.o vision_buffer.o MyODVDevice.o aria_action.o my_image.o MyResolver.o monitor_vel.o object_buffer.o UserTaskProlog.o aria_test.o ~/Robot/Aria/src/deliberative.pl ~/Robot/Aria/src/architectures.pl ~/Robot/Aria/src/associations.pl ~/Robot/Aria/src/beliefs.pl ~/Robot/Aria/src/belief_update.pl ~/Robot/Aria/src/choose_motivator.pl ~/Robot/Aria/src/desire_update.pl ~/Robot/Aria/src/goals.pl ~/Robot/Aria/src/intention_update.pl ~/Robot/Aria/src/motivator.pl ~/Robot/Aria/src/update_construction.pl -otest
/usr/bin/ld: cannot find -lAria
collect2: ld returned 1 exit status
gcc returned code 256
*** plld exit status 1
make: *** [test] Error 1
Is it pointing at some file (like test) or is it pointing at some other issue?
Josh
---
Hossein MIRI (Joshua) - Ph.D. Student
Neural, Emergent, and Agent Technologies Research Group (NEAT)
Department of Computer Science, University of Hull, Kingston-upon-Hull
East Riding of Yorkshire, HU6 7RX, United Kingdom
http://www.hmiri.com
http://hjoshua.squarespace.com
---
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080609/90d8b92a/attachment.html
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: not available
Url: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080609/90d8b92a/attachment.ksh
From wahyudi_firman at yahoo.com Thu Jun 5 19:14:25 2008
From: wahyudi_firman at yahoo.com (Firman Wahyudi)
Date: Thu, 5 Jun 2008 16:14:25 -0700 (PDT)
Subject: [Aria-users] robot pose value
Message-ID: <446290.45771.qm@web63813.mail.re1.yahoo.com>
Hello...
I try to get robot position while the robot wandering in the middle of room. I have use getX(), getY() or getPose function. but the return value always about 10 digit, like 1437462463, 1231243588. why the return value is too big?? Is there additional convertion to get the value become smaller?
I also do "standard calibration" at chapter 8. my robot get wrong about 4-5mm per 1000mm. and the degree about 0.5 per 90degree. and i think that's not a problem.
I also change "HasGyro 1 or 2". and try at gyroExample, but gyro always 0. I send the pic of gyroExample while HasGyro 1.
I am using P3DX and VS 2003 @ windows XP... and i also send my code...
thanks...
/*
This is the guts of the Turn action.
*/
ArActionDesired *ActionRobotMap::fire(ArActionDesired currentDesired)
{
// declare variable
ArPose poseSonar;
double delta;
double teta;
double th;
double deltaPRobot;
int sonarx, sonary;
// reset the actionDesired (must be done)
myDesired.reset();
// if the sonar is null we can't do anything, so deactivate
if (mySonar == NULL)
{
deactivate();
return NULL;
}
//ArPose* poseRobo = new ArPose(robot->getX(), robot->getY(), 0);
const ArRobotParams* robotParams = robot->getRobotParams();
int numsonar = robotParams->getNumSonar();
for (int i=0; igetSonarReading(i);
int range = robot->getSonarRange(i);
poseSonar = reading->getSensorPosition();
ArPose p = reading->getLocalPose();
ArPose pglobal = reading->getPose();
sonarx = ((ArRobotParams*)robotParams)->getSonarX(i);
sonary = ((ArRobotParams*)robotParams)->getSonarY(i);
// calculate robot distance from P
ArPose poseRobo = reading->getPoseTaken();
ArPose pRobo = robot->getPose();
// ArPose rpose = robot.getPose();
deltaPRobot = p.findDistanceTo(poseRobo);
delta = p.findDistanceTo(poseSonar);
teta = poseSonar.findAngleTo(p);
th = ((ArRobotParams*)robotParams)->getSonarTh(i);
teta = th-teta;
if ((((p.getX()!=0) && (p.getY()!=0)) && ((delta>=800)&&(delta<1200))) && ((teta>-10)&&(teta<10)))
{
// RobotMapper.updateMap
/*printf("sonar ke-%d \n", i);
printf("delta: %f \n", delta);
printf("th: %f \n", th);
printf("teta: %f \n", teta);
printf("posisi robot: (%d,%d) \n", p.getX(), p.getY());
printf("poseRobo: %d, %d \n", poseRobo.getX(), poseRobo.getY());
printf("pRobo: %d, %d \n", pRobo.getX(), pRobo.getY());
printf(rpose: %d, %d \n", rpose.getX(), rpose.getY());
printf("posisi awal: (%d,%d) \n", robot->getPose(), robot->getPose());
printf("start: (%d,%d) \n", robot->getX(), robot->getY());
printf("deltaPRobot: %f \n", deltaPRobot.getX(), deltaPRobot.getY());*/
mapper->updateMap(map, delta, teta, p);
//mapper->updateMapE2(map_e, map_occ, delta, teta, p);
}
}
// print matrix map ke file
//map->printMapToFile();
// turning robot
myDesired.setDeltaHeading(myTurnAmount * myTurning);
// return a pointer to the actionDesired, so resolver knows what to do
return &myDesired;
}
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080605/4d61f571/attachment-0001.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: gyroexample.jpg
Type: image/jpeg
Size: 158240 bytes
Desc: not available
Url : http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080605/4d61f571/attachment-0001.jpg
From ehady at students.tf.itb.ac.id Mon Jun 9 15:52:19 2008
From: ehady at students.tf.itb.ac.id (Eko Ahadyat)
Date: Tue, 10 Jun 2008 02:52:19 +0700 (WIT)
Subject: [Aria-users] "move to" inside "for(...)"
Message-ID: <4668.167.205.45.216.1213041139.squirrel@students.tf.itb.ac.id>
i want to give several tasks to the arm.
so i use "for(.....)" and array, just to make the program look simpler
i wrote :
for(int n=0;n<=3;n++)
{
printf("get the object...\n");
int getobject1[] = {0,t1o[n] , t2o[n], t3o[n], 0, t4o[n], -85};
for (int i=1; i<=ArP2Arm::NumJoints; i++)
{
joint = arm.getJoint(i);
arm.moveTo(i, getobject1[i], 0);
}
.
.
.
but the arm didn't response to this program...
why? what do you suggest?
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
From Jason.Chen at anu.edu.au Tue Jun 10 19:47:07 2008
From: Jason.Chen at anu.edu.au (Jason Chen)
Date: Wed, 11 Jun 2008 09:47:07 +1000
Subject: [Aria-users] TCM2 configuration and problems
Message-ID: <200806110947.07695.Jason.Chen@anu.edu.au>
Hi,
We have a pioneer P3-DX with a TCM2 module. The module seems to be returning
nonsensical readings for the compass value. If we run the demo program, the
pitch, roll and temperature values seem to make sense, ie when you tilt the
robot, these values change in a reasonable way. The compass and mag X,Y, and
Z values do not change in a reasonable way. The mag X,Y, and Z are always 0,
and the compass value changes in what seems to be a random way between
about -20 and +50 as you rotate the robot around its vertical axis. We are
running the robot in a typical indoor office environment not close, we
believe, to any large changing magnetic fields. We also mounted the TCM2 unit
on a 30cm high timber platform to keep it away from the robots
shell/chassis/motors.
We have looked at the documentation. There is the "TCM2 Compass, Inclinometer,
Magnetmometer, and Thermometer Manual", which seems to address the
installation of the TCM2 unit on the Pioneer 2, ie P2OS and LCD on
microcontroller ? We have also read the TCM2 manual from the manufacturers of
the unit ("TCM2 Manual"). There is an interesting part which talks about
driving in a circle while doing a user calibration. We tried to do this while
using the demo program for the calibration commands (using commands 4 and
then 6 after rotating the robot). However, it did not return any calibration
scores and did not change the strange compass values of the the mag X Y and Z
values from zero. Any ideas on how to correct this problem would be
appreciated
Cheers
Jason
From reed at mobilerobots.com Tue Jun 10 20:46:10 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 10 Jun 2008 20:46:10 -0400
Subject: [Aria-users] [Support] /usr/bin/ld: cannot find -lAria
In-Reply-To: <445800D8FD5D3D43878FD7D2842A3047898F7A@EXCL2VS2.adir.hull.ac.uk>
References: <445800D8FD5D3D43878FD7D2842A3047898F7A@EXCL2VS2.adir.hull.ac.uk>
Message-ID: <484F2052.9030904@mobilerobots.com>
Hello,
You probably need to add -L/usr/local/Aria/lib to your compilation flags
(before the -l flags).
Reed
Hossein Miri wrote:
> Hi Reed,
>
> When I run Make in my Aria directory (not the one that comes with the package - the one that has my Prolog and C++ programs) I get the bunch below. The first part is obviously my files, but what is the error message at the end?
>
> plld -c++ -I include -I/usr/local/Aria/include -I/usr/lib/pl-5.6.4/include/ -ldl -lpthread -lAria MyActionSonarConstantVel.o MyActionSonarAvoidSide.o MyActionSonarAvoidFrontNear.o MyActionSonarAvoidFrontFar.o MyActionSonarAvoidRear.o MyActionVisionConstantVel.o MyActionVisionAvoidFrontNear.o MyActionVisionAvoidFrontFar.o MyTrackObject.o MyHitObject.o vision_buffer.o MyODVDevice.o aria_action.o my_image.o MyResolver.o monitor_vel.o object_buffer.o UserTaskProlog.o aria_test.o ~/Robot/Aria/src/deliberative.pl ~/Robot/Aria/src/architectures.pl ~/Robot/Aria/src/associations.pl ~/Robot/Aria/src/beliefs.pl ~/Robot/Aria/src/belief_update.pl ~/Robot/Aria/src/choose_motivator.pl ~/Robot/Aria/src/desire_update.pl ~/Robot/Aria/src/goals.pl ~/Robot/Aria/src/intention_update.pl ~/Robot/Aria/src/motivator.pl ~/Robot/Aria/src/update_construction.pl -otest
>
> /usr/bin/ld: cannot find -lAria
> collect2: ld returned 1 exit status
> gcc returned code 256
> *** plld exit status 1
> make: *** [test] Error 1
>
>
> Is it pointing at some file (like test) or is it pointing at some other issue?
>
From reed at mobilerobots.com Tue Jun 10 20:47:56 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 10 Jun 2008 20:47:56 -0400
Subject: [Aria-users] About actsColorFollowingExample
In-Reply-To: <16072619.1095741211990343075.JavaMail.coremail@bj126app38.126.com>
References: <16072619.1095741211990343075.JavaMail.coremail@bj126app38.126.com>
Message-ID: <484F20BC.9040108@mobilerobots.com>
wjxnet_126 wrote:
> hello,
>
> i still want to ask why actsColorFollowingExample can not be run on our P3-AT, in the attachment is the error report. I can use Putty to log in P3-AT, then run vncserver, from vncviewer of a desktop, i can run acts demo, and the cameral transfers correct images. Thanks!
>
Hello,
Aria can't connect to ACTS Demo, only fully licensed ACTS. If you
need to purchase ACTS you can contact sales at mobilerobots.com.
Reed
From reed at mobilerobots.com Tue Jun 10 20:51:36 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 10 Jun 2008 20:51:36 -0400
Subject: [Aria-users] robot pose value
In-Reply-To: <446290.45771.qm@web63813.mail.re1.yahoo.com>
References: <446290.45771.qm@web63813.mail.re1.yahoo.com>
Message-ID: <484F2198.8080403@mobilerobots.com>
Firman Wahyudi wrote:
> Hello...
> I try to get robot position while the robot wandering in the middle of room. I have use getX(), getY() or getPose function. but the return value always about 10 digit, like 1437462463, 1231243588. why the return value is too big?? Is there additional convertion to get the value become smaller?
> I also do "standard calibration" at chapter 8. my robot get wrong about 4-5mm per 1000mm. and the degree about 0.5 per 90degree. and i think that's not a problem.
> I also change "HasGyro 1 or 2". and try at gyroExample, but gyro always 0. I send the pic of gyroExample while HasGyro 1.
> I am using P3DX and VS 2003 @ windows XP... and i also send my code...
>
Do you get those odd looking values when using %d when printing out the
values? They are all double floating point values, not integers, so you
should use %f.
Reed
> if ((((p.getX()!=0) && (p.getY()!=0)) && ((delta>=800)&&(delta<1200))) && ((teta>-10)&&(teta<10)))
> {
> // RobotMapper.updateMap
> /*printf("sonar ke-%d \n", i);
> printf("delta: %f \n", delta);
> printf("th: %f \n", th);
> printf("teta: %f \n", teta);
> printf("posisi robot: (%d,%d) \n", p.getX(), p.getY());
> printf("poseRobo: %d, %d \n", poseRobo.getX(), poseRobo.getY());
> printf("pRobo: %d, %d \n", pRobo.getX(), pRobo.getY());
> printf(rpose: %d, %d \n", rpose.getX(), rpose.getY());
> printf("posisi awal: (%d,%d) \n", robot->getPose(), robot->getPose());
> printf("start: (%d,%d) \n", robot->getX(), robot->getY());
> printf("deltaPRobot: %f \n", deltaPRobot.getX(), deltaPRobot.getY());*/
> mapper->updateMap(map, delta, teta, p);
> //mapper->updateMapE2(map_e, map_occ, delta, teta, p);
> }
> }
From reed at mobilerobots.com Tue Jun 10 20:53:27 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 10 Jun 2008 20:53:27 -0400
Subject: [Aria-users] robot pose value
In-Reply-To: <446290.45771.qm@web63813.mail.re1.yahoo.com>
References: <446290.45771.qm@web63813.mail.re1.yahoo.com>
Message-ID: <484F2207.7040706@mobilerobots.com>
Firman Wahyudi wrote:
> Hello...
> I try to get robot position while the robot wandering in the middle of room. I have use getX(), getY() or getPose function. but the return value always about 10 digit, like 1437462463, 1231243588. why the return value is too big?? Is there additional convertion to get the value become smaller?
> I also do "standard calibration" at chapter 8. my robot get wrong about 4-5mm per 1000mm. and the degree about 0.5 per 90degree. and i think that's not a problem.
> I also change "HasGyro 1 or 2". and try at gyroExample, but gyro always 0. I send the pic of gyroExample while HasGyro 1.
> I am using P3DX and VS 2003 @ windows XP... and i also send my code...
>
Regarding the gyro, do you know that your robot has a gyro? If you
don't know, ask support at mobilerobots.com, and include the robot's serial
number, so they can look it up.
Reed
From reed at mobilerobots.com Tue Jun 10 20:55:38 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 10 Jun 2008 20:55:38 -0400
Subject: [Aria-users] "move to" inside "for(...)"
In-Reply-To: <4668.167.205.45.216.1213041139.squirrel@students.tf.itb.ac.id>
References: <4668.167.205.45.216.1213041139.squirrel@students.tf.itb.ac.id>
Message-ID: <484F228A.2040200@mobilerobots.com>
Can you give more details? Does the arm move if you only do one loop
iteration, instead of 3? Do the contents of t1o etc. change at each
loop? Note that if you send two movement requests to the arm immediately
one after the other, the second one will cancel the first, it won't
store them.
Reed
From reed at mobilerobots.com Tue Jun 10 20:59:34 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 10 Jun 2008 20:59:34 -0400
Subject: [Aria-users] action not working -- was: wjxet_126@126.com:
Robots' initial positions when choosing "Create a new robot for each
connection" in MobileSim
In-Reply-To: <10808628.564901211906893497.JavaMail.coremail@bj126app38.126.com>
References: <483C26F9.8010700@mobilerobots.com>
<483C0663.1020705@mobilerobots.com>
<29251047.258051211620208581.JavaMail.coremail@bj126app20.126.com>
<29331480.934671211901319200.JavaMail.coremail@bj126app28.126.com>
<10808628.564901211906893497.JavaMail.coremail@bj126app38.126.com>
Message-ID: <484F2376.4010306@mobilerobots.com>
Hello,
Were you able to get this working? If not, can you give more details on
what this action is supposed to do, and what the robot does instead, and
also what other actions are being used?
Reed
wjxnet_126 at 126.com wrote:
> Anyhow, i write a action class to do some particular movement, but the robot, which is a server, does not move according to the class, the following is the codes
>
> class fileMode:public ArAction
> {
> public:
> fileMode(ArRobot *robot,ArTime *actionStart);
> virtual ~fileMode(void);
> virtual ArActionDesired *fire(ArActionDesired currentDesired);
> virtual void setRobot(ArRobot *robot);
> protected:
> ArActionDesired myDesired;
> ArRobot *myRobot;
> RobotTrace runnerTrace[RobotTraceNum];
> int num;
> ArTime *myStart;
> };
> // Constructor: Initialize the chase action
> fileMode::fileMode(ArRobot *robot, ArTime *actionStart) : ArAction("go")
> {
> // read robot's trace from a file to the struct runnerTrace
> }
> // Destructor
> fileMode::~fileMode(void) {}
>
> // The chase action
> ArActionDesired *fileMode::fire(ArActionDesired currentDesired)
> {
> myDesired.reset();
> myDesired.setHeading(runnerTrace[1].Th);
> myDesired.setVel(runnerTrace[1].Vel);
> return &myDesired;
> }
> void fileMode::setRobot(ArRobot *robot)
> {
> ArAction::setRobot(robot);
> }
>
>
> ArServerBase server;
> ArServerSimpleOpener simpleOpener(&parser);
> ArClientSwitchManager clientSwitchManager(&server, &parser);
> ....
>
> fileMode myFileMode(&robot,&actionStart);
> robot.addAction(&myFileMode,60);
>
> robot.runAsync(true);
> server.runAsync();
> clientSwitchManager.runAsync();
>
From sekmen at yahoo.com Tue Jun 10 22:39:44 2008
From: sekmen at yahoo.com (ali sekmen)
Date: Tue, 10 Jun 2008 19:39:44 -0700 (PDT)
Subject: [Aria-users] TCM2 configuration and problems
In-Reply-To: <200806110947.07695.Jason.Chen@anu.edu.au>
Message-ID: <284418.49305.qm@web36103.mail.mud.yahoo.com>
Try to change "3" to a higher number (the number of data you need to read) in the following line
myRobot->comInt(ArCommands::TCM2, 3);
Ali.
--- On Tue, 6/10/08, Jason Chen wrote:
> From: Jason Chen
> Subject: [Aria-users] TCM2 configuration and problems
> To: "aria"
> Date: Tuesday, June 10, 2008, 6:47 PM
> Hi,
> We have a pioneer P3-DX with a TCM2 module. The module
> seems to be returning
> nonsensical readings for the compass value. If we run the
> demo program, the
> pitch, roll and temperature values seem to make sense, ie
> when you tilt the
> robot, these values change in a reasonable way. The compass
> and mag X,Y, and
> Z values do not change in a reasonable way. The mag X,Y,
> and Z are always 0,
> and the compass value changes in what seems to be a random
> way between
> about -20 and +50 as you rotate the robot around its
> vertical axis. We are
> running the robot in a typical indoor office environment
> not close, we
> believe, to any large changing magnetic fields. We also
> mounted the TCM2 unit
> on a 30cm high timber platform to keep it away from the
> robots
> shell/chassis/motors.
>
> We have looked at the documentation. There is the
> "TCM2 Compass, Inclinometer,
> Magnetmometer, and Thermometer Manual", which seems to
> address the
> installation of the TCM2 unit on the Pioneer 2, ie P2OS and
> LCD on
> microcontroller ? We have also read the TCM2 manual from
> the manufacturers of
> the unit ("TCM2 Manual"). There is an interesting
> part which talks about
> driving in a circle while doing a user calibration. We
> tried to do this while
> using the demo program for the calibration commands (using
> commands 4 and
> then 6 after rotating the robot). However, it did not
> return any calibration
> scores and did not change the strange compass values of the
> the mag X Y and Z
> values from zero. Any ideas on how to correct this problem
> would be
> appreciated
> Cheers
> Jason
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail
> to:
>
> aria-users-leave at lists.mobilerobots.com
From Jason.Chen at anu.edu.au Tue Jun 10 23:37:01 2008
From: Jason.Chen at anu.edu.au (Jason Chen)
Date: Wed, 11 Jun 2008 13:37:01 +1000
Subject: [Aria-users] TCM2 configuration and problems
In-Reply-To: <284418.49305.qm@web36103.mail.mud.yahoo.com>
References: <284418.49305.qm@web36103.mail.mud.yahoo.com>
Message-ID: <200806111337.02210.Jason.Chen@anu.edu.au>
Hi Ali,
thanks for the advice. However it is not that we are reading too little data.
We can see at a lower level the data coming across from the microcontroller
and it all seems to be there, its just that the compass and mag values seem
erroneous
Jason
On Wednesday 11 June 2008 12:39, ali sekmen wrote:
> Try to change "3" to a higher number (the number of data you need to read)
> in the following line
>
> myRobot->comInt(ArCommands::TCM2, 3);
>
> Ali.
>
> --- On Tue, 6/10/08, Jason Chen wrote:
> > From: Jason Chen
> > Subject: [Aria-users] TCM2 configuration and problems
> > To: "aria"
> > Date: Tuesday, June 10, 2008, 6:47 PM
> > Hi,
> > We have a pioneer P3-DX with a TCM2 module. The module
> > seems to be returning
> > nonsensical readings for the compass value. If we run the
> > demo program, the
> > pitch, roll and temperature values seem to make sense, ie
> > when you tilt the
> > robot, these values change in a reasonable way. The compass
> > and mag X,Y, and
> > Z values do not change in a reasonable way. The mag X,Y,
> > and Z are always 0,
> > and the compass value changes in what seems to be a random
> > way between
> > about -20 and +50 as you rotate the robot around its
> > vertical axis. We are
> > running the robot in a typical indoor office environment
> > not close, we
> > believe, to any large changing magnetic fields. We also
> > mounted the TCM2 unit
> > on a 30cm high timber platform to keep it away from the
> > robots
> > shell/chassis/motors.
> >
> > We have looked at the documentation. There is the
> > "TCM2 Compass, Inclinometer,
> > Magnetmometer, and Thermometer Manual", which seems to
> > address the
> > installation of the TCM2 unit on the Pioneer 2, ie P2OS and
> > LCD on
> > microcontroller ? We have also read the TCM2 manual from
> > the manufacturers of
> > the unit ("TCM2 Manual"). There is an interesting
> > part which talks about
> > driving in a circle while doing a user calibration. We
> > tried to do this while
> > using the demo program for the calibration commands (using
> > commands 4 and
> > then 6 after rotating the robot). However, it did not
> > return any calibration
> > scores and did not change the strange compass values of the
> > the mag X Y and Z
> > values from zero. Any ideas on how to correct this problem
> > would be
> > appreciated
> > Cheers
> > Jason
> >
> > _______________________________________________
> > Aria-users mailing list
> > Aria-users at lists.mobilerobots.com
> > http://lists.mobilerobots.com/mailman/listinfo/aria-users
> >
> > To unsubscribe visit the above webpage or send an e-mail
> > to:
> >
> > aria-users-leave at lists.mobilerobots.com
From yuema.bit at gmail.com Wed Jun 11 00:54:04 2008
From: yuema.bit at gmail.com (Yue Ma)
Date: Wed, 11 Jun 2008 12:54:04 +0800
Subject: [Aria-users] submission
Message-ID: <332f16ad0806102154t5fb68226w360afd3b74efd1f3@mail.gmail.com>
as above
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080611/57cf3ae7/attachment.html
From H.Miri at 2007.hull.ac.uk Wed Jun 11 05:42:50 2008
From: H.Miri at 2007.hull.ac.uk (Hossein Miri)
Date: Wed, 11 Jun 2008 10:42:50 +0100
Subject: [Aria-users] [Support] /usr/bin/ld: cannot find -lAria
References: <445800D8FD5D3D43878FD7D2842A3047898F7A@EXCL2VS2.adir.hull.ac.uk>
<484F2052.9030904@mobilerobots.com>
Message-ID: <445800D8FD5D3D43878FD7D2842A3047898F81@EXCL2VS2.adir.hull.ac.uk>
Hi Reed,
My makefile seems OK. Where exactly do you think I should stick the line you have suggested?
Cheers,
# my own makefile for my ARIA and Prolog and C++ programs
# set up variables
CXX=g++
ARIA_PATH=/usr/local/Aria
LIBS=-ldl -lpthread -lAria
ARIA_INCLUDE=$(ARIA_PATH)/include
PROLOG_INCLUDE=/usr/lib/pl-5.6.4/include/
LIBS = -ldl -lpthread -lAria
CPPFLAGS = -I include -I$(ARIA_INCLUDE) -I$(PROLOG_INCLUDE)
VPATH = src src/actions
EXE_FILE=test
CFILES=\
MyActionSonarConstantVel.cpp \
MyActionSonarAvoidSide.cpp \
MyActionSonarAvoidFrontNear.cpp \
MyActionSonarAvoidFrontFar.cpp \
MyActionSonarAvoidRear.cpp \
MyActionVisionConstantVel.cpp \
MyActionVisionAvoidFrontNear.cpp \
MyActionVisionAvoidFrontFar.cpp \
MyTrackObject.cpp \
MyHitObject.cpp \
vision_buffer.cpp \
MyODVDevice.cpp \
aria_action.cpp \
my_image.cpp \
MyResolver.cpp \
monitor_vel.cpp \
object_buffer.cpp \
UserTaskProlog.cpp \
aria_test.cpp
OFILES=\
MyActionSonarConstantVel.o \
MyActionSonarAvoidSide.o \
MyActionSonarAvoidFrontNear.o \
MyActionSonarAvoidFrontFar.o \
MyActionSonarAvoidRear.o \
MyActionVisionConstantVel.o \
MyActionVisionAvoidFrontNear.o \
MyActionVisionAvoidFrontFar.o \
MyTrackObject.o \
MyHitObject.o \
vision_buffer.o \
MyODVDevice.o \
aria_action.o \
my_image.o \
MyResolver.o \
monitor_vel.o \
object_buffer.o \
UserTaskProlog.o \
aria_test.o
PLFILES=\
~/Robot/Aria/src/deliberative.pl \
~/Robot/Aria/src/architectures.pl \
~/Robot/Aria/src/associations.pl \
~/Robot/Aria/src/beliefs.pl \
~/Robot/Aria/src/belief_update.pl \
~/Robot/Aria/src/choose_motivator.pl \
~/Robot/Aria/src/desire_update.pl \
~/Robot/Aria/src/goals.pl \
~/Robot/Aria/src/intention_update.pl \
~/Robot/Aria/src/motivator.pl \
~/Robot/Aria/src/update_construction.pl
# target rules and dependences
$(EXE_FILE):$(OFILES)
# auto dependences
# must update dependancies
depend: $(CFILES)
$(CXX) -M $(CPPFLAGS) $^ > $@
include depend
# create object files
%.o:%.cpp
$(CXX) $(CPPFLAGS) -c $<
#$(EXE_FILE):
# $(CXX) $(CPPFLAGS) $(LIBS) -o$(EXE_FILE) $(OFILES)
$(EXE_FILE):
plld -c++ $(CPPFLAGS) $(LIBS) $(OFILES) $(PLFILES) -o$(EXE_FILE)
---
Hossein MIRI (Joshua) - Ph.D. Student
Neural, Emergent, and Agent Technologies Research Group (NEAT)
Department of Computer Science, University of Hull, Kingston-upon-Hull
East Riding of Yorkshire, HU6 7RX, United Kingdom
http://www.hmiri.com
http://hjoshua.squarespace.com
---
________________________________
Hello,
You probably need to add -L/usr/local/Aria/lib to your compilation flags
(before the -l flags).
Reed
________________________________
> Hi Reed,
>
> When I run Make in my Aria directory (not the one that comes with the package - the one that has my Prolog and C++ programs) I get the bunch below. The first part is obviously my files, but what is the error message at the end?
>
> plld -c++ -I include -I/usr/local/Aria/include -I/usr/lib/pl-5.6.4/include/ -ldl -lpthread -lAria MyActionSonarConstantVel.o MyActionSonarAvoidSide.o MyActionSonarAvoidFrontNear.o MyActionSonarAvoidFrontFar.o MyActionSonarAvoidRear.o MyActionVisionConstantVel.o MyActionVisionAvoidFrontNear.o MyActionVisionAvoidFrontFar.o MyTrackObject.o MyHitObject.o vision_buffer.o MyODVDevice.o aria_action.o my_image.o MyResolver.o monitor_vel.o object_buffer.o UserTaskProlog.o aria_test.o ~/Robot/Aria/src/deliberative.pl ~/Robot/Aria/src/architectures.pl ~/Robot/Aria/src/associations.pl ~/Robot/Aria/src/beliefs.pl ~/Robot/Aria/src/belief_update.pl ~/Robot/Aria/src/choose_motivator.pl ~/Robot/Aria/src/desire_update.pl ~/Robot/Aria/src/goals.pl ~/Robot/Aria/src/intention_update.pl ~/Robot/Aria/src/motivator.pl ~/Robot/Aria/src/update_construction.pl -otest
>
> /usr/bin/ld: cannot find -lAria
> collect2: ld returned 1 exit status
> gcc returned code 256
> *** plld exit status 1
> make: *** [test] Error 1
>
>
> Is it pointing at some file (like test) or is it pointing at some other issue?
_______________________________________________
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080611/d1b67260/attachment-0001.html
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: not available
Url: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080611/d1b67260/attachment-0001.ksh
From H.Miri at 2007.hull.ac.uk Wed Jun 11 10:59:30 2008
From: H.Miri at 2007.hull.ac.uk (Hossein Miri)
Date: Wed, 11 Jun 2008 15:59:30 +0100
Subject: [Aria-users] compat - libstdc++
References: <445800D8FD5D3D43878FD7D2842A3047898F7A@EXCL2VS2.adir.hull.ac.uk>
<484F2052.9030904@mobilerobots.com>
Message-ID: <445800D8FD5D3D43878FD7D2842A3047898F82@EXCL2VS2.adir.hull.ac.uk>
Hi all,
As you may know, Fedora 4 ships and uses GCC 4.0. Some applications that were compiled with earlier versions of GCC will require compatibility libraries. Does anybody know why I get the error message below when I try:
$ yum install compat-libstdc++-33
$ yum install compat-libstdc++-296
$ yum install compat-gcc-32
$ yum install compat-gcc-32-c++
Error Message: Cannot find a valid baseurl for repo: updates-released
.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080611/c588d071/attachment.html
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: not available
Url: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080611/c588d071/attachment.ksh
From martin at mobilerobots.com Wed Jun 11 11:49:50 2008
From: martin at mobilerobots.com (Martin Ekendahl)
Date: Wed, 11 Jun 2008 11:49:50 -0400
Subject: [Aria-users] compat - libstdc++
In-Reply-To: <445800D8FD5D3D43878FD7D2842A3047898F82@EXCL2VS2.adir.hull.ac.uk>
References: <445800D8FD5D3D43878FD7D2842A3047898F7A@EXCL2VS2.adir.hull.ac.uk> <484F2052.9030904@mobilerobots.com>
<445800D8FD5D3D43878FD7D2842A3047898F82@EXCL2VS2.adir.hull.ac.uk>
Message-ID: <484FF41E.90009@mobilerobots.com>
Hossein Miri wrote:
>
> Error Message: Cannot find a valid baseurl for repo: updates-released
>
You might want to try some of the suggestions listed in this thread:
http://www.linuxquestions.org/questions/linux-software-2/yum-error-cannot-find-a-valid-baseurl-for-repo-478600/
Mostly is seems you need to uncomment baseurl from *.repo files in
/etc/yum.repos.d/ and double check your network settings.
-Martin
--
Martin Ekendahl
martin at mobilerobots.com
IT Manager
MobileRobots Inc.
Amherst, NH
From aaron123 at ksu.edu Wed Jun 11 16:48:20 2008
From: aaron123 at ksu.edu (aaron123 at ksu.edu)
Date: Wed, 11 Jun 2008 15:48:20 -0500
Subject: [Aria-users] ACTS using uncompressed video data
Message-ID: <1213217300.48503a1475f69@webmail.ksu.edu>
Hello everyone. I've read the ACTS manual and there doesn't appear to
be a way to run ACTS using uncompressed data coming from the camera but
I would really like to do this even if it means using a slower frame
rate. The other alternative is to use ACTS to process images that I
pass to it although I see no documentation on this type of feature
which leads me to believe that it isn't possible. Does anyone have any
ideas on a good approach/option?
Thanks,
Aaron Chleborad
From H.Miri at 2007.hull.ac.uk Thu Jun 12 10:37:35 2008
From: H.Miri at 2007.hull.ac.uk (Hossein Miri)
Date: Thu, 12 Jun 2008 15:37:35 +0100
Subject: [Aria-users] compat - libstdc++
References: <445800D8FD5D3D43878FD7D2842A3047898F7A@EXCL2VS2.adir.hull.ac.uk> <484F2052.9030904@mobilerobots.com>
<445800D8FD5D3D43878FD7D2842A3047898F82@EXCL2VS2.adir.hull.ac.uk>
<484FF41E.90009@mobilerobots.com>
Message-ID: <445800D8FD5D3D43878FD7D2842A3047898F84@EXCL2VS2.adir.hull.ac.uk>
Hi Martin,
Thanks for the webpage. I, also, uncommented the line you suggested in all .repo files, but I got a new error message:
Cannot open/read repomd.xml file for repository: updates-released
failure: repodata/repomd.xml from updates-released: [Errno 256] No more mirrors to try.
Any ideas on how to resolve this one?
Cheers,
________________________________
Hossein Miri wrote:
>
> Error Message: Cannot find a valid baseurl for repo: updates-released
>
You might want to try some of the suggestions listed in this thread:
http://www.linuxquestions.org/questions/linux-software-2/yum-error-cannot-find-a-valid-baseurl-for-repo-478600/
Mostly is seems you need to uncomment baseurl from *.repo files in
/etc/yum.repos.d/ and double check your network settings.
-Martin
--
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080612/4d254d14/attachment.html
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: not available
Url: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080612/4d254d14/attachment.ksh
From martin at mobilerobots.com Thu Jun 12 10:54:31 2008
From: martin at mobilerobots.com (Martin Ekendahl)
Date: Thu, 12 Jun 2008 10:54:31 -0400
Subject: [Aria-users] compat - libstdc++
In-Reply-To: <445800D8FD5D3D43878FD7D2842A3047898F84@EXCL2VS2.adir.hull.ac.uk>
References: <445800D8FD5D3D43878FD7D2842A3047898F7A@EXCL2VS2.adir.hull.ac.uk> <484F2052.9030904@mobilerobots.com>
<445800D8FD5D3D43878FD7D2842A3047898F82@EXCL2VS2.adir.hull.ac.uk>
<484FF41E.90009@mobilerobots.com>
<445800D8FD5D3D43878FD7D2842A3047898F84@EXCL2VS2.adir.hull.ac.uk>
Message-ID: <485138A7.6050009@mobilerobots.com>
Hossein Miri wrote:
> Hi Martin,
>
> Thanks for the webpage. I, also, uncommented the line you suggested in
> all .repo files, but I got a new error message:
>
> Cannot open/read repomd.xml file for repository: updates-released
> failure: repodata/repomd.xml from updates-released: [Errno 256] No more
> mirrors to try.
>
> Any ideas on how to resolve this one?
>
> Cheers,
>
>
> ------------------------------------------------------------------------
>
> Hossein Miri wrote:
>
> >
> > Error Message: Cannot find a valid baseurl for repo: updates-released
> >
>
> You might want to try some of the suggestions listed in this thread:
>
> http://www.linuxquestions.org/questions/linux-software-2/yum-error-cannot-find-a-valid-baseurl-for-repo-478600/
>
> Mostly is seems you need to uncomment baseurl from *.repo files in
> /etc/yum.repos.d/ and double check your network settings.
>
> -Martin
>
My memory on yum is a little hazy, but it sounds like either yum is
timing out on that mirror or the mirror you have is no longer reachable.
Try adding one that is close to you with appropriate bandwidth.
From this list: http://mirrors.fedoraproject.org/publiclist
ftp://ftp.mirrorservice.org/pub/Fedora/releases/
looks promising for your location. You'll have to double check the
syntax in your current conf file, I don't have a fedora box to check here.
You might also try disabling some repositories while trying to install
that software
yum --disablerepo=legacy-base --disablerepo=legacy-updates install "package"
Also try throwing that error message into google, someone else must have
run into the problem who has more fedora expertise than me, they might
have some other ideas =)
-Martin
---
Martin Ekendahl
martin at mobilerobots.com
IT Manager
MobileRobots Inc.
Amherst, NH
From albertocorralesgarcia at gmail.com Thu Jun 12 14:56:09 2008
From: albertocorralesgarcia at gmail.com (Alberto Corrales Garcia)
Date: Thu, 12 Jun 2008 20:56:09 +0200
Subject: [Aria-users] video server
In-Reply-To: <4803CDA4.1010101@mobilerobots.com>
References: <1672c7d80804101152k3856e1e4u5ae50252291a252d@mail.gmail.com>
<4803CDA4.1010101@mobilerobots.com>
Message-ID: <1672c7d80806121156h69f4b29at760364f4846315d7@mail.gmail.com>
hello,
I'm seeing getVideoExample() and I saw it uses synchronous communication,
but Is it the best option or asynchronous communicacion is better to
implement a video server-client?
thanks for your help!
2008/4/14 Reed Hedges :
>
> Hi Alberto,
>
> What SAV server does is compress the image to a JPEG and send that using
> ArNetworking. You can use libjpeg to do that on Linux (it's a standard
> library, you can install it from RedHat or use apt-get on Debian).
>
> To transfer the images using ArNetworking, on the server set up a packet
> handler to respond to "getPictureCam1" requests. In each reply packet put a
> pair of 16-bit integers containing image width and height in pixels using
> uByte2ToBuf(), then the JPEG data using dataToBuf(). The ArNetworking
> example getVideoExample.cpp requests images using getPictureCam1 and this
> packet format. MobileEyes does too, so you can use that as your client if
> you just want to see the image (you could also then use ArConfig to store
> any parameters and use ArServerHandlerConfig to let MobileEyes modify the
> values in ArConfig too.)
>
> See the ArNetworking reference documentation and the other examples if
> you're not familiar with it. Sorry there isn't as much info about
> ArNetworking as ARIA, we're still working on that.
>
> Reed
>
>
> --
>
> Reed Hedges
> Developer Software Support
> MobileRobots Inc.
>
> Downloads, Documentation, FAQ, and more:
>
>
> aria-users Mailing List Archive:
>
>
> aria-users Mailing List Subscribe and Unsubscribe Instructions:
>
>
>
--
-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=
Alberto Corrales Garc?a
Coordinador DotNetClub Albacete
blog: http://labloguera.net/blogs/alberto/
web: http://albertocorralesgarcia.googlepages.com
PFC: http://pfc-albertocorralesgarcia.blogspot.com/
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080612/3bdbcd46/attachment-0001.html
From reed at mobilerobots.com Fri Jun 13 11:11:25 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 13 Jun 2008 11:11:25 -0400
Subject: [Aria-users] video server
In-Reply-To: <1672c7d80806121156h69f4b29at760364f4846315d7@mail.gmail.com>
References: <1672c7d80804101152k3856e1e4u5ae50252291a252d@mail.gmail.com> <4803CDA4.1010101@mobilerobots.com>
<1672c7d80806121156h69f4b29at760364f4846315d7@mail.gmail.com>
Message-ID: <48528E1D.7050902@mobilerobots.com>
Hello,
I don't think I fully understand what you mean, could you explain in more
detail? What do you mean by synchronous communication in getVideoExample? Is
there a specific problem you're encountering or anticipate?
ArNetworking's client-side components run asynchronously from the main thread of
the program (though each callback for data recieved from the server is
synchronous with respect to the other callbacks.)
Reed
Alberto Corrales Garcia wrote:
> hello,
>
> I'm seeing getVideoExample() and I saw it uses synchronous
> communication, but Is it the best option or asynchronous communicacion
> is better to implement a video server-client?
>
> thanks for your help!
>
> 2008/4/14 Reed Hedges >:
>
>
> Hi Alberto,
>
> What SAV server does is compress the image to a JPEG and send that
> using ArNetworking. You can use libjpeg to do that on Linux (it's a
> standard library, you can install it from RedHat or use apt-get on
> Debian).
>
> To transfer the images using ArNetworking, on the server set up a
> packet handler to respond to "getPictureCam1" requests. In each
> reply packet put a pair of 16-bit integers containing image width
> and height in pixels using uByte2ToBuf(), then the JPEG data using
> dataToBuf(). The ArNetworking example getVideoExample.cpp requests
> images using getPictureCam1 and this packet format. MobileEyes does
> too, so you can use that as your client if you just want to see the
> image (you could also then use ArConfig to store any parameters and
> use ArServerHandlerConfig to let MobileEyes modify the values in
> ArConfig too.)
>
> See the ArNetworking reference documentation and the other examples
> if you're not familiar with it. Sorry there isn't as much info about
> ArNetworking as ARIA, we're still working on that.
>
> Reed
>
>
> --
>
> Reed Hedges
> Developer Software Support
> MobileRobots Inc.
>
> Downloads, Documentation, FAQ, and more:
>
>
> aria-users Mailing List Archive:
>
>
> aria-users Mailing List Subscribe and Unsubscribe Instructions:
>
>
>
>
>
> --
>
> -=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=
> Alberto Corrales Garc?a
> Coordinador DotNetClub Albacete
> blog: http://labloguera.net/blogs/alberto/
> web: http://albertocorralesgarcia.googlepages.com
> PFC: http://pfc-albertocorralesgarcia.blogspot.com/
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail to:
>
> aria-users-leave at lists.mobilerobots.com
From reed at mobilerobots.com Fri Jun 13 11:17:32 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 13 Jun 2008 11:17:32 -0400
Subject: [Aria-users] TCM2 configuration and problems
In-Reply-To: <200806111337.02210.Jason.Chen@anu.edu.au>
References: <284418.49305.qm@web36103.mail.mud.yahoo.com>
<200806111337.02210.Jason.Chen@anu.edu.au>
Message-ID: <48528F8C.9090600@mobilerobots.com>
Hi Jason, what version of ARIA are you using? (See version.txt or Changes.txt).
My guess is that the calibration is not working right, have you tried it again?
You can try doing it by hand, instead of by rotating the robot, and see if
that makes a difference.
Reed
Jason Chen wrote:
> Hi Ali,
> thanks for the advice. However it is not that we are reading too little data.
> We can see at a lower level the data coming across from the microcontroller
> and it all seems to be there, its just that the compass and mag values seem
> erroneous
> Jason
>
> On Wednesday 11 June 2008 12:39, ali sekmen wrote:
>> Try to change "3" to a higher number (the number of data you need to read)
>> in the following line
>>
>> myRobot->comInt(ArCommands::TCM2, 3);
>>
>> Ali.
>>
>> --- On Tue, 6/10/08, Jason Chen wrote:
>>> From: Jason Chen
>>> Subject: [Aria-users] TCM2 configuration and problems
>>> To: "aria"
>>> Date: Tuesday, June 10, 2008, 6:47 PM
>>> Hi,
>>> We have a pioneer P3-DX with a TCM2 module. The module
>>> seems to be returning
>>> nonsensical readings for the compass value. If we run the
>>> demo program, the
>>> pitch, roll and temperature values seem to make sense, ie
>>> when you tilt the
>>> robot, these values change in a reasonable way. The compass
>>> and mag X,Y, and
>>> Z values do not change in a reasonable way. The mag X,Y,
>>> and Z are always 0,
>>> and the compass value changes in what seems to be a random
>>> way between
>>> about -20 and +50 as you rotate the robot around its
>>> vertical axis. We are
>>> running the robot in a typical indoor office environment
>>> not close, we
>>> believe, to any large changing magnetic fields. We also
>>> mounted the TCM2 unit
>>> on a 30cm high timber platform to keep it away from the
>>> robots
>>> shell/chassis/motors.
>>>
>>> We have looked at the documentation. There is the
>>> "TCM2 Compass, Inclinometer,
>>> Magnetmometer, and Thermometer Manual", which seems to
>>> address the
>>> installation of the TCM2 unit on the Pioneer 2, ie P2OS and
>>> LCD on
>>> microcontroller ? We have also read the TCM2 manual from
>>> the manufacturers of
>>> the unit ("TCM2 Manual"). There is an interesting
>>> part which talks about
>>> driving in a circle while doing a user calibration. We
>>> tried to do this while
>>> using the demo program for the calibration commands (using
>>> commands 4 and
>>> then 6 after rotating the robot). However, it did not
>>> return any calibration
>>> scores and did not change the strange compass values of the
>>> the mag X Y and Z
>>> values from zero. Any ideas on how to correct this problem
>>> would be
>>> appreciated
>>> Cheers
>>> Jason
>>>
>>> _______________________________________________
>>> Aria-users mailing list
>>> Aria-users at lists.mobilerobots.com
>>> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>>>
>>> To unsubscribe visit the above webpage or send an e-mail
>>> to:
>>>
>>> aria-users-leave at lists.mobilerobots.com
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail to:
>
> aria-users-leave at lists.mobilerobots.com
>
From Jason.Chen at anu.edu.au Sun Jun 15 20:00:30 2008
From: Jason.Chen at anu.edu.au (Jason Chen)
Date: Mon, 16 Jun 2008 10:00:30 +1000
Subject: [Aria-users] TCM2 configuration and problems
In-Reply-To: <48528F8C.9090600@mobilerobots.com>
References: <284418.49305.qm@web36103.mail.mud.yahoo.com>
<200806111337.02210.Jason.Chen@anu.edu.au>
<48528F8C.9090600@mobilerobots.com>
Message-ID: <200806161000.30468.Jason.Chen@anu.edu.au>
Hi Reed,
thanks for the reply. I'm going to try to test it out of the robot, ie with
the little battery powered cable thing that came with the TCM package. I'm
just waiting on getting access to a pc with a serial port.
BTW, I saw in the documentation that there is a .EXE file to communicate with
it more easily than with minicom, but I can't find this software, either on
the CD's that were sent out with the robot or on your web site. Could you
advise where I can get it
Cheers
Jason
On Saturday 14 June 2008 01:17, Reed Hedges wrote:
> Hi Jason, what version of ARIA are you using? (See version.txt or
> Changes.txt).
>
> My guess is that the calibration is not working right, have you tried it
> again? You can try doing it by hand, instead of by rotating the robot, and
> see if that makes a difference.
>
> Reed
>
> Jason Chen wrote:
> > Hi Ali,
> > thanks for the advice. However it is not that we are reading too little
> > data. We can see at a lower level the data coming across from the
> > microcontroller and it all seems to be there, its just that the compass
> > and mag values seem erroneous
> > Jason
> >
> > On Wednesday 11 June 2008 12:39, ali sekmen wrote:
> >> Try to change "3" to a higher number (the number of data you need to
> >> read) in the following line
> >>
> >> myRobot->comInt(ArCommands::TCM2, 3);
> >>
> >> Ali.
> >>
> >> --- On Tue, 6/10/08, Jason Chen wrote:
> >>> From: Jason Chen
> >>> Subject: [Aria-users] TCM2 configuration and problems
> >>> To: "aria"
> >>> Date: Tuesday, June 10, 2008, 6:47 PM
> >>> Hi,
> >>> We have a pioneer P3-DX with a TCM2 module. The module
> >>> seems to be returning
> >>> nonsensical readings for the compass value. If we run the
> >>> demo program, the
> >>> pitch, roll and temperature values seem to make sense, ie
> >>> when you tilt the
> >>> robot, these values change in a reasonable way. The compass
> >>> and mag X,Y, and
> >>> Z values do not change in a reasonable way. The mag X,Y,
> >>> and Z are always 0,
> >>> and the compass value changes in what seems to be a random
> >>> way between
> >>> about -20 and +50 as you rotate the robot around its
> >>> vertical axis. We are
> >>> running the robot in a typical indoor office environment
> >>> not close, we
> >>> believe, to any large changing magnetic fields. We also
> >>> mounted the TCM2 unit
> >>> on a 30cm high timber platform to keep it away from the
> >>> robots
> >>> shell/chassis/motors.
> >>>
> >>> We have looked at the documentation. There is the
> >>> "TCM2 Compass, Inclinometer,
> >>> Magnetmometer, and Thermometer Manual", which seems to
> >>> address the
> >>> installation of the TCM2 unit on the Pioneer 2, ie P2OS and
> >>> LCD on
> >>> microcontroller ? We have also read the TCM2 manual from
> >>> the manufacturers of
> >>> the unit ("TCM2 Manual"). There is an interesting
> >>> part which talks about
> >>> driving in a circle while doing a user calibration. We
> >>> tried to do this while
> >>> using the demo program for the calibration commands (using
> >>> commands 4 and
> >>> then 6 after rotating the robot). However, it did not
> >>> return any calibration
> >>> scores and did not change the strange compass values of the
> >>> the mag X Y and Z
> >>> values from zero. Any ideas on how to correct this problem
> >>> would be
> >>> appreciated
> >>> Cheers
> >>> Jason
> >>>
> >>> _______________________________________________
> >>> Aria-users mailing list
> >>> Aria-users at lists.mobilerobots.com
> >>> http://lists.mobilerobots.com/mailman/listinfo/aria-users
> >>>
> >>> To unsubscribe visit the above webpage or send an e-mail
> >>> to:
> >>>
> >>> aria-users-leave at lists.mobilerobots.com
> >
> > _______________________________________________
> > Aria-users mailing list
> > Aria-users at lists.mobilerobots.com
> > http://lists.mobilerobots.com/mailman/listinfo/aria-users
> >
> > To unsubscribe visit the above webpage or send an e-mail to:
> >
> > aria-users-leave at lists.mobilerobots.com
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail to:
>
> aria-users-leave at lists.mobilerobots.com
From reed at mobilerobots.com Mon Jun 16 08:13:05 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Mon, 16 Jun 2008 08:13:05 -0400
Subject: [Aria-users] TCM2 configuration and problems
In-Reply-To: <200806161818.31575.Jason.Chen@anu.edu.au>
References: <284418.49305.qm@web36103.mail.mud.yahoo.com>
<200806111337.02210.Jason.Chen@anu.edu.au>
<48528F8C.9090600@mobilerobots.com>
<200806161818.31575.Jason.Chen@anu.edu.au>
Message-ID: <485658D1.407@mobilerobots.com>
Great!
The microcontroller should be sending those commands to the compass, but I will
look into it. (Or, bad connection between compass and microcontroller before?)
Reed
Jason Chen wrote:
> Hi Reed,
> just to let you know, I managed to solve the problem by directly connecting to
> the compass unit with the supplied cable and a pc with a serial port. I used
> the cc command to clear previous (I suspect) badly performed calibration and
> then the compass readings made sense. I needed to enable the magnetometer
> output (em=e) so that the compass unit was indeed sending this data to the
> microcontroller to put in the TCM2 packet
> Cheers
> Jason
>
>
> On Saturday 14 June 2008 01:17, you wrote:
>> Hi Jason, what version of ARIA are you using? (See version.txt or
>> Changes.txt).
>>
>> My guess is that the calibration is not working right, have you tried it
>> again? You can try doing it by hand, instead of by rotating the robot, and
>> see if that makes a difference.
>>
>> Reed
>>
>> Jason Chen wrote:
>>> Hi Ali,
>>> thanks for the advice. However it is not that we are reading too little
>>> data. We can see at a lower level the data coming across from the
>>> microcontroller and it all seems to be there, its just that the compass
>>> and mag values seem erroneous
>>> Jason
From albertocorralesgarcia at gmail.com Tue Jun 17 04:40:23 2008
From: albertocorralesgarcia at gmail.com (Alberto Corrales Garcia)
Date: Tue, 17 Jun 2008 10:40:23 +0200
Subject: [Aria-users] video server
In-Reply-To: <48528E1D.7050902@mobilerobots.com>
References: <1672c7d80804101152k3856e1e4u5ae50252291a252d@mail.gmail.com>
<4803CDA4.1010101@mobilerobots.com>
<1672c7d80806121156h69f4b29at760364f4846315d7@mail.gmail.com>
<48528E1D.7050902@mobilerobots.com>
Message-ID: <1672c7d80806170140p1ec7ec52n86e95521d185363d@mail.gmail.com>
yes, I think if the source and the receiver would be synchronize sendig
frames to see the video correctly. I saw the class ArHybridForwarderVideo,
which has some methods to manage that subject, but I don't know how it work
and I didn't see any example. Could you tell me about that class?
Regards,
Alberto
2008/6/13 Reed Hedges :
>
>
> Hello,
>
> I don't think I fully understand what you mean, could you explain in more
> detail? What do you mean by synchronous communication in getVideoExample?
> Is
> there a specific problem you're encountering or anticipate?
>
> ArNetworking's client-side components run asynchronously from the main
> thread of
> the program (though each callback for data recieved from the server is
> synchronous with respect to the other callbacks.)
>
>
> Reed
>
>
> Alberto Corrales Garcia wrote:
> > hello,
> >
> > I'm seeing getVideoExample() and I saw it uses synchronous
> > communication, but Is it the best option or asynchronous communicacion
> > is better to implement a video server-client?
> >
> > thanks for your help!
> >
> > 2008/4/14 Reed Hedges > >:
> >
> >
> > Hi Alberto,
> >
> > What SAV server does is compress the image to a JPEG and send that
> > using ArNetworking. You can use libjpeg to do that on Linux (it's a
> > standard library, you can install it from RedHat or use apt-get on
> > Debian).
> >
> > To transfer the images using ArNetworking, on the server set up a
> > packet handler to respond to "getPictureCam1" requests. In each
> > reply packet put a pair of 16-bit integers containing image width
> > and height in pixels using uByte2ToBuf(), then the JPEG data using
> > dataToBuf(). The ArNetworking example getVideoExample.cpp requests
> > images using getPictureCam1 and this packet format. MobileEyes does
> > too, so you can use that as your client if you just want to see the
> > image (you could also then use ArConfig to store any parameters and
> > use ArServerHandlerConfig to let MobileEyes modify the values in
> > ArConfig too.)
> >
> > See the ArNetworking reference documentation and the other examples
> > if you're not familiar with it. Sorry there isn't as much info about
> > ArNetworking as ARIA, we're still working on that.
> >
> > Reed
> >
> >
> > --
> >
> > Reed Hedges
> > Developer Software Support
> > MobileRobots Inc.
> >
> > Downloads, Documentation, FAQ, and more:
> >
> >
> > aria-users Mailing List Archive:
> >
> >
> > aria-users Mailing List Subscribe and Unsubscribe Instructions:
> >
> >
> >
> >
> >
> > --
> >
> > -=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=
> > Alberto Corrales Garc?a
> > Coordinador DotNetClub Albacete
> > blog: http://labloguera.net/blogs/alberto/
> > web: http://albertocorralesgarcia.googlepages.com
> > PFC: http://pfc-albertocorralesgarcia.blogspot.com/
> > -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> >
> >
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > Aria-users mailing list
> > Aria-users at lists.mobilerobots.com
> > http://lists.mobilerobots.com/mailman/listinfo/aria-users
> >
> > To unsubscribe visit the above webpage or send an e-mail to:
> >
> > aria-users-leave at lists.mobilerobots.com
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail to:
>
> aria-users-leave at lists.mobilerobots.com
--
-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=
Alberto Corrales Garc?a
Coordinador DotNetClub Albacete
blog: http://labloguera.net/blogs/alberto/
web: http://albertocorralesgarcia.googlepages.com
PFC: http://pfc-albertocorralesgarcia.blogspot.com/
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080617/6df1dc7c/attachment-0001.html
From reed at mobilerobots.com Tue Jun 17 16:56:08 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 17 Jun 2008 16:56:08 -0400
Subject: [Aria-users] video server
In-Reply-To: <1672c7d80806170140p1ec7ec52n86e95521d185363d@mail.gmail.com>
References: <1672c7d80804101152k3856e1e4u5ae50252291a252d@mail.gmail.com>
<4803CDA4.1010101@mobilerobots.com>
<1672c7d80806121156h69f4b29at760364f4846315d7@mail.gmail.com>
<48528E1D.7050902@mobilerobots.com>
<1672c7d80806170140p1ec7ec52n86e95521d185363d@mail.gmail.com>
Message-ID: <485824E8.6050004@mobilerobots.com>
Alberto Corrales Garcia wrote:
> yes, I think if the source and the receiver would be synchronize sendig
> frames to see the video correctly. I saw the class
> ArHybridForwarderVideo, which has some methods to manage that subject,
> but I don't know how it work and I didn't see any example. Could you
> tell me about that class?
Hi Alberto,
The client will receive each packet from the server (source) in turn, at
whatever rate it requests. For example if you call:
client.addHandler("getPictureCam1", &handlerFunctor);
client.request("getPictureCam1", 30);
Then the client will send the getPictureCam1 request to the server 30 times a
second. And the client's callback function will be called serially for each
response packet is received (one thread).
When the server receives a request it may then read an image from the
framegrabber, or it may have an image stored from an asynchronous thread... I
don't remember how exactly it works right now but will look at some of the code
to see if I can determine that.
However, in either case the server won't get "backed up"-- there is no queue of
images waiting to be sent, it's always the most recent one it has. (How your
client program handles the images is up to you though.) It is possible to
"miss" a potentially unique image, but you can minimize this by using a request
rate faster than the framegrabber's own rate of generating images (you will get
some duplicate images in that case).
So the client is synchronized with the server by the request-response mechanism,
but this is may be asynchronous in a sense from how the framegrabber or other
source is providing images or compared to the ideal sequence of unique images
you could be getting.
ArHybridForwarderVideo is a class used to forward ArNetworking video requests
and data obtained from another ArNetworkingServer. For example if you start
ACTS or SAVServer, it provides video images on port 7070 using ArNetworking. If
you then start ARNL's guiServer, it contains an ArHybridForwarderVideo object,
which grabs the images from ACTS or SAVServer, and turns around and provides
them on the same port as the rest of its services (robot info, path planning,
map, config, etc.). Then you can connect with MobileEyes and get all the
services, including video images, from the same server and port.
Does this make sense? Let me know if there are any details you need me to find
out.
Reed
From reed at mobilerobots.com Tue Jun 17 17:06:07 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 17 Jun 2008 17:06:07 -0400
Subject: [Aria-users] ACTS using uncompressed video data
In-Reply-To: <1213217300.48503a1475f69@webmail.ksu.edu>
References: <1213217300.48503a1475f69@webmail.ksu.edu>
Message-ID: <4858273F.1060002@mobilerobots.com>
Hi Aaron,
What do you mean by uncompressed data from the camera? What kind of
framegrabber are you using?
There may be some framegrabber-specific options available... check the Video
Parameters dialog, or run ACTS on a command line with the -P option to see a
list of them. Then you can run it with one of those options set on a command
line by including the -s option. See Table 1 in chapter 2 of the ACTS manual.
You can have ACTS work on a single image you provide (Load Image in the menu),
but not a stream of images.
If simply selecting a slower rate helps, you can change that too... though I
think that just changes the rate at which ACTS acquires the images from the
framegrabber.
Reed
aaron123 at ksu.edu wrote:
> Hello everyone. I've read the ACTS manual and there doesn't appear to
> be a way to run ACTS using uncompressed data coming from the camera but
> I would really like to do this even if it means using a slower frame
> rate. The other alternative is to use ACTS to process images that I
> pass to it although I see no documentation on this type of feature
> which leads me to believe that it isn't possible. Does anyone have any
> ideas on a good approach/option?
From blomba at unione.com.br Wed Jun 18 12:14:42 2008
From: blomba at unione.com.br (Bruno Teixeira Lomba)
Date: Wed, 18 Jun 2008 13:14:42 -0300
Subject: [Aria-users] SLAM - HELP
Message-ID: <200806181314.AA94830932@unione.com.br>
> Hi...
>
> I?m starting my studies in SLAM (Simultaneous Localization and
> Mapping), and I?d like to take some questions, before beginning to
> implement the PeopleBot.
>
> - There is ready some library/source code about SLAM???
>
> I want to use Monte Carlo Localization algorithm (MCL) with the Bayes
> rule, which at that researched is the most efficient ...
>
> - As the robot creates the map to navigate the simulator
> (MobileEyes)???
>
> - Where do I include the SLAM in existing codes of ARIA, ARNL ...???
>
> I'm still a little confused about the use of these tools ...
>
> Somebody could help me???
>
> Then my email contact: blomba at unione.com.br
>
> Thank you very much
>
> Hugs
>
> Bruno Lomba
From mshaker at lincoln.ac.uk Thu Jun 19 09:30:47 2008
From: mshaker at lincoln.ac.uk (Marwan Shaker)
Date: Thu, 19 Jun 2008 14:30:47 +0100
Subject: [Aria-users] Looking for help
Message-ID: <139F6E4BC4C585478FA84D02D90CC1CA0DB674AB@aexcmb02.network.uni>
Dear sir I want to ask . If I can call aria ( not using the main
function ( not in console ) but under WinMain, the problem is that when
we define the main as below
int main(int argc, char** argv) , and the command
ArSimpleConnector conn(&argc, argv) use the argc, and argv, but in the
WinMain there is no such arguments there is different which they are
HINSTANCE hInstance,HINSTANCE hPrevInstance, LPSTR lpCmdLine, int
nCmdShow
How can I use then in the simpleconnecter
ArSimpleConnector conn(&argc, argv)
With best wishes
From albertocorralesgarcia at gmail.com Thu Jun 19 13:00:40 2008
From: albertocorralesgarcia at gmail.com (Alberto Corrales Garcia)
Date: Thu, 19 Jun 2008 19:00:40 +0200
Subject: [Aria-users] video server
In-Reply-To: <485824E8.6050004@mobilerobots.com>
References: <1672c7d80804101152k3856e1e4u5ae50252291a252d@mail.gmail.com>
<4803CDA4.1010101@mobilerobots.com>
<1672c7d80806121156h69f4b29at760364f4846315d7@mail.gmail.com>
<48528E1D.7050902@mobilerobots.com>
<1672c7d80806170140p1ec7ec52n86e95521d185363d@mail.gmail.com>
<485824E8.6050004@mobilerobots.com>
Message-ID: <1672c7d80806191000o71d54dsa881eaa8a6af3772@mail.gmail.com>
Hello and thank you for your help,
I'm learning about Aria and ArNetworking but I'm a little confusing, because
some subjects are different to the traditional C++ programming.
I would be very pleased if you could send me code for a server video or tell
me how I can respond to the video requests.
Regards
2008/6/17 Reed Hedges :
> Alberto Corrales Garcia wrote:
> > yes, I think if the source and the receiver would be synchronize sendig
> > frames to see the video correctly. I saw the class
> > ArHybridForwarderVideo, which has some methods to manage that subject,
> > but I don't know how it work and I didn't see any example. Could you
> > tell me about that class?
>
> Hi Alberto,
>
> The client will receive each packet from the server (source) in turn, at
> whatever rate it requests. For example if you call:
>
> client.addHandler("getPictureCam1", &handlerFunctor);
> client.request("getPictureCam1", 30);
>
> Then the client will send the getPictureCam1 request to the server 30 times
> a
> second. And the client's callback function will be called serially for each
> response packet is received (one thread).
>
> When the server receives a request it may then read an image from the
> framegrabber, or it may have an image stored from an asynchronous thread...
> I
> don't remember how exactly it works right now but will look at some of the
> code
> to see if I can determine that.
>
> However, in either case the server won't get "backed up"-- there is no
> queue of
> images waiting to be sent, it's always the most recent one it has. (How
> your
> client program handles the images is up to you though.) It is possible to
> "miss" a potentially unique image, but you can minimize this by using a
> request
> rate faster than the framegrabber's own rate of generating images (you will
> get
> some duplicate images in that case).
>
> So the client is synchronized with the server by the request-response
> mechanism,
> but this is may be asynchronous in a sense from how the framegrabber or
> other
> source is providing images or compared to the ideal sequence of unique
> images
> you could be getting.
>
> ArHybridForwarderVideo is a class used to forward ArNetworking video
> requests
> and data obtained from another ArNetworkingServer. For example if you
> start
> ACTS or SAVServer, it provides video images on port 7070 using
> ArNetworking. If
> you then start ARNL's guiServer, it contains an ArHybridForwarderVideo
> object,
> which grabs the images from ACTS or SAVServer, and turns around and
> provides
> them on the same port as the rest of its services (robot info, path
> planning,
> map, config, etc.). Then you can connect with MobileEyes and get all the
> services, including video images, from the same server and port.
>
> Does this make sense? Let me know if there are any details you need me to
> find
> out.
>
> Reed
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail to:
>
> aria-users-leave at lists.mobilerobots.com
>
--
-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=
Alberto Corrales Garc?a
Coordinador DotNetClub Albacete
blog: http://labloguera.net/blogs/alberto/
web: http://albertocorralesgarcia.googlepages.com
PFC: http://pfc-albertocorralesgarcia.blogspot.com/
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080619/72edac77/attachment.html
From blomba at unione.com.br Thu Jun 19 13:11:22 2008
From: blomba at unione.com.br (Bruno Teixeira Lomba)
Date: Thu, 19 Jun 2008 14:11:22 -0300
Subject: [Aria-users] SLAM - created map
Message-ID: <200806191411.AA1053556860@unione.com.br>
Hello.
I am learning to use SLAM (Simultaneous Localization and Mapping) for using on PeopleBot ...
How the map is created when I do get the robot navigate?
I could not find it anywhere ...
I saw something in a message posted in the forum who said something like:
Connect to this server with MobileEyes
Go to Tools->Map Creation->Start Mapping
Give the map a name and hit okay
Drive the robot around your space (see docs/Mapping.txt)
Go to Tools->Map Creation->Stop Mapping
Start up Mapper3
Go to File->Open on Robot
Select the .2d you created
Create a .map
Go to File->Save on Robot
With MobileEyes go to Tools->Robot Config
Change the Map entry to be your new .map
Press okay and your new map should become the map used
But I could not find the option Tools-> Map Creation ...
Could you help me?
Thank you.
Bruno Lomba
From Jason.Chen at anu.edu.au Thu Jun 19 22:27:16 2008
From: Jason.Chen at anu.edu.au (Jason Chen)
Date: Fri, 20 Jun 2008 12:27:16 +1000
Subject: [Aria-users] bump sensors sticking
In-Reply-To: <4858273F.1060002@mobilerobots.com>
References: <1213217300.48503a1475f69@webmail.ksu.edu>
<4858273F.1060002@mobilerobots.com>
Message-ID: <200806201227.16582.Jason.Chen@anu.edu.au>
Hi Reed,
I've been having problems with bump sensors sticking sometimes after a
collision. That is, they stay jammed-in after the robot has collided with an
object and then moved away. As you can imagine, this causes no end of
problems for any control software. I was wondering whether other people have
had the same problem. I was thinking maybe a solution is stiffer springs,
since the robot with batteries and laser etc is pretty heavy and has no
problem compressing the bump sensors as they come as standard.
Jason
From Jason.Chen at anu.edu.au Thu Jun 19 22:27:16 2008
From: Jason.Chen at anu.edu.au (Jason Chen)
Date: Fri, 20 Jun 2008 12:27:16 +1000
Subject: [Aria-users] bump sensors sticking
In-Reply-To: <4858273F.1060002@mobilerobots.com>
References: <1213217300.48503a1475f69@webmail.ksu.edu>
<4858273F.1060002@mobilerobots.com>
Message-ID: <200806201227.16582.Jason.Chen@anu.edu.au>
Hi Reed,
I've been having problems with bump sensors sticking sometimes after a
collision. That is, they stay jammed-in after the robot has collided with an
object and then moved away. As you can imagine, this causes no end of
problems for any control software. I was wondering whether other people have
had the same problem. I was thinking maybe a solution is stiffer springs,
since the robot with batteries and laser etc is pretty heavy and has no
problem compressing the bump sensors as they come as standard.
Jason
From reed at mobilerobots.com Fri Jun 20 08:53:45 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 20 Jun 2008 08:53:45 -0400
Subject: [Aria-users] SLAM - created map
In-Reply-To: <200806191411.AA1053556860@unione.com.br>
References: <200806191411.AA1053556860@unione.com.br>
Message-ID: <485BA859.1020306@mobilerobots.com>
Hi Bruno,
The first step in making a map is to capture a log of scanning laser data as you
drive the robot around an environment. This scan log can be generated by
ARNL's "guiServer" example program, or by the "sickLogger" program. If you run
guiServer on the robot, you can connect to it using the MobileEyes program (e.g.
over the wireless network), and you should have the "Map Creation" options in
the Tools menu of MobileEyes.
Or you can use sickLogger instead of guiServer -- sickLogger only logs laser
data, it doesn't have any other features or services.
After you have using Stop Mapping, or ended sickLogger, you will have a log file
ending in ".2d". This file can be loaded into the Mapper3 program, which will
use a SLAM algorithm to generate a real map from that data.
Reed
Bruno Teixeira Lomba wrote:
> Hello.
>
> I am learning to use SLAM (Simultaneous Localization and Mapping) for using on PeopleBot ...
>
> How the map is created when I do get the robot navigate?
>
> I could not find it anywhere ...
>
> I saw something in a message posted in the forum who said something like:
>
> Connect to this server with MobileEyes
> Go to Tools->Map Creation->Start Mapping
> Give the map a name and hit okay
> Drive the robot around your space (see docs/Mapping.txt)
> Go to Tools->Map Creation->Stop Mapping
> Start up Mapper3
> Go to File->Open on Robot
> Select the .2d you created
> Create a .map
> Go to File->Save on Robot
> With MobileEyes go to Tools->Robot Config
> Change the Map entry to be your new .map
> Press okay and your new map should become the map used
>
> But I could not find the option Tools-> Map Creation ...
>
> Could you help me?
>
> Thank you.
>
> Bruno Lomba
From reed at mobilerobots.com Fri Jun 20 08:58:51 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 20 Jun 2008 08:58:51 -0400
Subject: [Aria-users] Looking for help
In-Reply-To: <139F6E4BC4C585478FA84D02D90CC1CA0DB674AB@aexcmb02.network.uni>
References: <139F6E4BC4C585478FA84D02D90CC1CA0DB674AB@aexcmb02.network.uni>
Message-ID: <485BA98B.7070004@mobilerobots.com>
Hi Marwan,
I don't know very much about using MFC, but if you can convert lpCmdLine into an
array of strings (char**), or can otherwise get such an array from MFC, then you
can pass those to ArSimpleConnector. Or, you can create an ArArgumentBuilder
object, and pass that to ArSimpleConnector instead of (argc, argv).
ArArgumentBuilder has methods for adding argument strings one by one, or parsing
a string at spaces into seperate argument strings, etc.
Reed
Marwan Shaker wrote:
> Dear sir I want to ask . If I can call aria ( not using the main
> function ( not in console ) but under WinMain, the problem is that when
> we define the main as below
> int main(int argc, char** argv) , and the command
>
> ArSimpleConnector conn(&argc, argv) use the argc, and argv, but in the
> WinMain there is no such arguments there is different which they are
>
> HINSTANCE hInstance,HINSTANCE hPrevInstance, LPSTR lpCmdLine, int
> nCmdShow
>
> How can I use then in the simpleconnecter
>
> ArSimpleConnector conn(&argc, argv)
>
> With best wishes
From reed at mobilerobots.com Fri Jun 20 09:33:53 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 20 Jun 2008 09:33:53 -0400
Subject: [Aria-users] SLAM - HELP
In-Reply-To: <200806181314.AA94830932@unione.com.br>
References: <200806181314.AA94830932@unione.com.br>
Message-ID: <485BB1C1.7000604@mobilerobots.com>
Hi Bruno,
We don't provide any code for doing SLAM. The Mapper3 program uses a SLAM
algorithm to convert laser scan data into a useful map, but it's not available
as a library or source code.
ARNL provides two main features: localization and path planning/following. You
can use these features together or seperately. The localization is based on MCL.
If you want to develop your own localization or SLAM code, it seems like you
would get the most use out of using ARIA directly. You can use ARIA to do the
robot motion, access robot and sensor data etc., while also doing localization.
There are a few things in ARIA that might be useful: ArAsyncTask for running a
parallel thread (alongside the robot control) (but you need to manage the
concurrency, lock objects being used in other threads, etc.), or user tasks (or
sensor interpretation tasks) for doing incremental work each ArRobot iteration
(every 100ms). When you have calculated a new position estimate for the robot,
you can use ArRobot::moveTo() to set ArRobot's pose to your new position. (This
is how ARNL's localization task sets a new corrected position).
It's then also possible to use ARNL's path planning and following task
(ArPathPlanningTask), but not use its localization (and do you own localization
instead.) Note that if you are using only ARIA, you can just use the ARIA
library from the ARIA package. If you are using both ARNL and ARIA together, use
the ARNL and AriaForArnl and ArNetworkingForArnl libraries that are included
along with ARNL (all compatible with each other).
Read the ARIA API Reference manual for more about how to use ARIA, and see its
example programs.
Does this help?
Reed
Bruno Teixeira Lomba wrote:
>> Hi...
>>
>> I?m starting my studies in SLAM (Simultaneous Localization and
>> Mapping), and I?d like to take some questions, before beginning to
>> implement the PeopleBot.
>>
>> - There is ready some library/source code about SLAM???
>>
>> I want to use Monte Carlo Localization algorithm (MCL) with the Bayes
>> rule, which at that researched is the most efficient ...
>>
>> - As the robot creates the map to navigate the simulator
>> (MobileEyes)???
>>
>> - Where do I include the SLAM in existing codes of ARIA, ARNL ...???
>>
>> I'm still a little confused about the use of these tools ...
>>
>> Somebody could help me???
>>
>> Then my email contact: blomba at unione.com.br
>>
>> Thank you very much
>>
>> Hugs
>>
>> Bruno Lomba
>
> ____________________________
From aaron123 at ksu.edu Tue Jun 24 13:51:14 2008
From: aaron123 at ksu.edu (aaron123 at ksu.edu)
Date: Tue, 24 Jun 2008 12:51:14 -0500
Subject: [Aria-users] Java 1.6 wrapper update
Message-ID: <1214329874.48613412dc0f8@webmail.ksu.edu>
Has anyone created a recent JNI wrapper for Aria? Our project is using
Java 1.6 but using the current JNI wrapper causes quirky (and
frustrating) errors. I've attempted recompiling the project on Windows
using Visual Studio and SWIG but the resulting files don't work
properly. As I am not an expert with SWIG, this is making progress on
the project difficult. Much of the code written uses Java 1.6
constructs so reverting back to Java 1.4.2 isn't an option.
The main problem we are having is when dealing with the ArSick data. We
require the raw data from the laser and the only way to obtain it is
using the getRawReadingsAsVector method. From there, we using the
ArSensorReadingVector.get method to obtain an ArSensorReading from
which we get the range value. This works for about 30 seconds then for
some reason a JVM call is made... AriaJavaJNI.delete_ArSerialConnection
.. which kills the serial connection and crashes the app. I was able
to find this problem using the -verbose:jni -Xcheck:jni VM options when
running the program. I have confirmed that when compiling and running
the laser code in isolation using Java 1.4.2 it runs fine.
Assistance would be much appreciated.
Thanks,
Aaron
From albertocorralesgarcia at gmail.com Tue Jun 24 14:35:35 2008
From: albertocorralesgarcia at gmail.com (Alberto Corrales Garcia)
Date: Tue, 24 Jun 2008 20:35:35 +0200
Subject: [Aria-users] Mapper3
Message-ID: <1672c7d80806241135p3218237p628652e619420635@mail.gmail.com>
hello,
I have a map with Autocad format and I have to build manually the same map
on the tool Mapper3-basic (because I don't have any laser) to access with
Aria. But, what's the equivalent to a real meter in this tool?
Regards
--
-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=
Alberto Corrales Garc?a
Coordinador DotNetClub Albacete
blog: http://labloguera.net/blogs/alberto/
web: http://albertocorralesgarcia.googlepages.com
PFC: http://pfc-albertocorralesgarcia.blogspot.com/
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mobilerobots.com/pipermail/aria-users/attachments/20080624/20c143d5/attachment.html
From mjohnson at ihmc.us Tue Jun 24 14:55:18 2008
From: mjohnson at ihmc.us (Matthew Johnson)
Date: Tue, 24 Jun 2008 13:55:18 -0500
Subject: [Aria-users] Java 1.6 wrapper update
In-Reply-To: <1214329874.48613412dc0f8@webmail.ksu.edu>
References: <1214329874.48613412dc0f8@webmail.ksu.edu>
Message-ID: <000601c8d62b$d87a7160$1a00640a@mjdell>
We had a lot of trouble with the Java version of Aria as well. Which
version of Aria are you using?
We have it working under Aria2.5.1 and jdk1.5.0_09. If you think that would
work I can send it to you.
Matt
-----Original Message-----
From: aria-users-bounces at lists.mobilerobots.com
[mailto:aria-users-bounces at lists.mobilerobots.com] On Behalf Of
aaron123 at ksu.edu
Sent: Tuesday, June 24, 2008 12:51 PM
To: aria-users at mobilerobots.com
Subject: [Aria-users] Java 1.6 wrapper update
Has anyone created a recent JNI wrapper for Aria? Our project is using
Java 1.6 but using the current JNI wrapper causes quirky (and
frustrating) errors. I've attempted recompiling the project on Windows
using Visual Studio and SWIG but the resulting files don't work
properly. As I am not an expert with SWIG, this is making progress on
the project difficult. Much of the code written uses Java 1.6
constructs so reverting back to Java 1.4.2 isn't an option.
The main problem we are having is when dealing with the ArSick data. We
require the raw data from the laser and the only way to obtain it is
using the getRawReadingsAsVector method. From there, we using the
ArSensorReadingVector.get method to obtain an ArSensorReading from
which we get the range value. This works for about 30 seconds then for
some reason a JVM call is made... AriaJavaJNI.delete_ArSerialConnection
.. which kills the serial connection and crashes the app. I was able
to find this problem using the -verbose:jni -Xcheck:jni VM options when
running the program. I have confirmed that when compiling and running
the laser code in isolation using Java 1.4.2 it runs fine.
Assistance would be much appreciated.
Thanks,
Aaron
_______________________________________________
Aria-users mailing list
Aria-users at lists.mobilerobots.com
http://lists.mobilerobots.com/mailman/listinfo/aria-users
To unsubscribe visit the above webpage or send an e-mail to:
aria-users-leave at lists.mobilerobots.com
From aaron123 at ksu.edu Tue Jun 24 15:28:50 2008
From: aaron123 at ksu.edu (aaron123 at ksu.edu)
Date: Tue, 24 Jun 2008 14:28:50 -0500
Subject: [Aria-users] Java 1.6 wrapper update
In-Reply-To: <000601c8d62b$d87a7160$1a00640a@mjdell>
References: <1214329874.48613412dc0f8@webmail.ksu.edu>
<000601c8d62b$d87a7160$1a00640a@mjdell>
Message-ID: <1214335730.48614af2d582e@webmail.ksu.edu>
Thanks for your response Matt. We are using Aria 2.5.1 and we'd
definitely be interested in your updated version. If you could send
it we'd be very appreciative.
Thanks,
Aaron
Quoting Matthew Johnson :
> We had a lot of trouble with the Java version of Aria as well. Which
> version of Aria are you using?
>
> We have it working under Aria2.5.1 and jdk1.5.0_09. If you think
> that would
> work I can send it to you.
>
> Matt
>
> -----Original Message-----
> From: aria-users-bounces at lists.mobilerobots.com
> [mailto:aria-users-bounces at lists.mobilerobots.com] On Behalf Of
> aaron123 at ksu.edu
> Sent: Tuesday, June 24, 2008 12:51 PM
> To: aria-users at mobilerobots.com
> Subject: [Aria-users] Java 1.6 wrapper update
>
> Has anyone created a recent JNI wrapper for Aria? Our project is
> using
> Java 1.6 but using the current JNI wrapper causes quirky (and
> frustrating) errors. I've attempted recompiling the project on
> Windows
> using Visual Studio and SWIG but the resulting files don't work
> properly. As I am not an expert with SWIG, this is making progress
> on
> the project difficult. Much of the code written uses Java 1.6
> constructs so reverting back to Java 1.4.2 isn't an option.
>
> The main problem we are having is when dealing with the ArSick data.
> We
> require the raw data from the laser and the only way to obtain it is
> using the getRawReadingsAsVector method. From there, we using the
> ArSensorReadingVector.get method to obtain an ArSensorReading from
> which we get the range value. This works for about 30 seconds then
> for
> some reason a JVM call is made...
> AriaJavaJNI.delete_ArSerialConnection
> .. which kills the serial connection and crashes the app. I was able
> to find this problem using the -verbose:jni -Xcheck:jni VM options
> when
> running the program. I have confirmed that when compiling and
> running
> the laser code in isolation using Java 1.4.2 it runs fine.
>
> Assistance would be much appreciated.
>
>
> Thanks,
> Aaron
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail to:
>
> aria-users-leave at lists.mobilerobots.com
>
>
>
From reed at mobilerobots.com Tue Jun 24 17:39:06 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Tue, 24 Jun 2008 17:39:06 -0400
Subject: [Aria-users] Mapper3
In-Reply-To: <1672c7d80806241135p3218237p628652e619420635@mail.gmail.com>
References: <1672c7d80806241135p3218237p628652e619420635@mail.gmail.com>
Message-ID: <4861697A.4050507@mobilerobots.com>
Hi Alberto,
Inside the map file, millimeters are used.
In Mapper3, the grid lines (Map menu -> Grid) are each meter.
The values shown in the bottom status bar of the window and when you
double click on an object with the select tool are millimeters -- or you
can change them in View->Units menu item (to meters or feet/inches).
Reed
Alberto Corrales Garcia wrote:
> hello,
>
> I have a map with Autocad format and I have to build manually the same map
> on the tool Mapper3-basic (because I don't have any laser) to access with
> Aria. But, what's the equivalent to a real meter in this tool?
>
> Regards
>
>
From mshaker at lincoln.ac.uk Wed Jun 25 05:15:47 2008
From: mshaker at lincoln.ac.uk (Marwan Shaker)
Date: Wed, 25 Jun 2008 10:15:47 +0100
Subject: [Aria-users] Asking question
Message-ID: <139F6E4BC4C585478FA84D02D90CC1CA0DB674AE@aexcmb02.network.uni>
Hi all
I heard that there is wrapper for ARIA from c++ to c#, can any one tell
me where can I find this wrapper ( because it will solve my most problem
with c++, by the way is there a software for the aria IN C#, or not
With best wishes
From reed at mobilerobots.com Wed Jun 25 07:49:12 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Wed, 25 Jun 2008 07:49:12 -0400
Subject: [Aria-users] Asking question - ARIA with C#
In-Reply-To: <139F6E4BC4C585478FA84D02D90CC1CA0DB674AE@aexcmb02.network.uni>
References: <139F6E4BC4C585478FA84D02D90CC1CA0DB674AE@aexcmb02.network.uni>
Message-ID: <486230B8.6010802@mobilerobots.com>
Hi Marwan,
There is no C# wrapper library for ARIA, just Java and Python.
Supposedly, you can link a managed or unmanaged C++ library to C# in the CLR,
but I don't know a lot about that. I think you would need to recompile ARIA
with CLR enabled, and also may need to write some special declarations in C# for
the C++ methods you want to use (P/Invoke)?
If anyone else on the list knows more about it, do you have any tips or links to
resources?
Reed
Marwan Shaker wrote:
> Hi all
>
> I heard that there is wrapper for ARIA from c++ to c#, can any one tell
> me where can I find this wrapper ( because it will solve my most problem
> with c++, by the way is there a software for the aria IN C#, or not
>
> With best wishes
From blomba at unione.com.br Wed Jun 25 09:01:44 2008
From: blomba at unione.com.br (Bruno Teixeira Lomba)
Date: Wed, 25 Jun 2008 10:01:44 -0300
Subject: [Aria-users] SONARNL - convert created map
Message-ID: <200806251001.AA193003744@unione.com.br>
Hi,
I?m developing SLAM for the PeopleBot..
The robot that I am using, doesn?t have Laser...
SONAR only once, so I?m using SONARNL ..
Is it true that it doesn?t create maps?
How do the part of Mapping of SLAM then?
Could you help me?
If he makes mapping, as can I see the map created?
How convert the data from sonar to the format of output of ARNL?
Thank you for help.
Bruno Lomba
From aaron123 at ksu.edu Thu Jun 26 14:38:07 2008
From: aaron123 at ksu.edu (aaron123 at ksu.edu)
Date: Thu, 26 Jun 2008 13:38:07 -0500
Subject: [Aria-users] Java 1.6 wrapper update - progress update
In-Reply-To: <000601c8d6f8$e5c6e5b0$1a00640a@mjdell>
References: <1214329874.48613412dc0f8@webmail.ksu.edu>
<000601c8d62b$d87a7160$1a00640a@mjdell>
<1214335730.48614af2d582e@webmail.ksu.edu>
<000301c8d641$877a32d0$3b00640a@mjdell>
<1214343638.486169d61e440@webmail.ksu.edu>
<000001c8d66c$5f481270$6401a8c0@mjdell>
<1214419912.486293c88603b@webmail.ksu.edu>
<000601c8d6f8$e5c6e5b0$1a00640a@mjdell>
Message-ID: <1214505487.4863e20f98459@webmail.ksu.edu>
Hi Matt. Here is the final update on the situation. I have confirmed
that the files you sent us work great. I have also been able to create
a swig wrapper for Aria 2.5.1 and Java 1.6 (Using Visual Studio 2005
Pro) that runs well in debug mode. I have further found that for some
reason when the program is run in normal mode that the garbage
collector seems to trigger some odd behavior in the swig code... The
VM closes the serial connection to the laser on its own (I've seen
similar behavior where the VM kills ArSimpleConnector, ArArgumentParser
and ArArgumentBuilder). What's even stranger is that when I add code
that attempts to re-establish the serial connection when it is closed
the problem goes away completely and the code runs fine. I have found
that using these VM options at run time "-Xcheck:jni -verbose:gc
-verbose:jni ", I have been able to isolate where and what errant JNI
calls are made and can add robustness code which eliminates the
crashes. I've even been able to run different versions of the Swig
wrapper with newer versions of Java in this way.
Thanks a bunch for all you help!
Aaron Chleborad
Kansas State University
Department of Computing and Information Sciences
Research Assistant
http://macr.cis.ksu.edu/
Quoting Matthew Johnson :
> By changing the namespace in the jar, we were required to rebuild the
> dlls.
> This means that the jar I provided should not work with other dlls
> and the
> dlls I provided will not work with other jars. Make sure that the
> jars and
> dlls are the first (or preferably only) ones in the path and
> classpath.
> Unsatisfied link error is the likely result of using our jar with an
> old
> dll.
>
> Matt
>
> -----Original Message-----
> From: aaron123 at ksu.edu [mailto:aaron123 at ksu.edu]
> Sent: Wednesday, June 25, 2008 1:52 PM
> To: Matthew Johnson
> Subject: RE: [Aria-users] Java 1.6 wrapper update
>
> I spotted that someone had placed some Aria files in the system32
> folder
> and Java was using them instead of the new files. So, the app runs
> but
> still blows up with a pure virtual function call. I'm going to try
> compiling the app with Java 1.5 instead of 1.6...
>
> A
>
> Quoting Matthew Johnson :
>
> > Then you'll probably need these to go with the jars.
> >
> > Let me know how it works for you.
> >
> > Matt
> >
> > -----Original Message-----
> > From: aaron123 at ksu.edu [mailto:aaron123 at ksu.edu]
> > Sent: Tuesday, June 24, 2008 4:41 PM
> > To: Matthew Johnson
> > Subject: RE: [Aria-users] Java 1.6 wrapper update
> >
> > We are using Windows XP Pro.
> >
> > Aaron
> >
> >
> > Quoting Matthew Johnson :
> >
> > >
> > > Aaron,
> > >
> > > I hope this works for you. It took a lot of effort to repackage
> > > Aria. They
> > > did not package their code which made repackaging a problem for
> > JNI.
> > > Are
> > > you using Windows or Linux? You may need the rebuilt libraries
> as
> > > well.
> > >
> > > Matt
> > >
> > >
> > >
> > > -----Original Message-----
> > > From: aaron123 at ksu.edu [mailto:aaron123 at ksu.edu]
> > > Sent: Tuesday, June 24, 2008 2:29 PM
> > > To: Matthew Johnson
> > > Cc: aria-users at mobilerobots.com
> > > Subject: RE: [Aria-users] Java 1.6 wrapper update
> > >
> > > Thanks for your response Matt. We are using Aria 2.5.1 and we'd
> > > definitely be interested in your updated version. If you could
> > send
> > > it we'd be very appreciative.
> > >
> > > Thanks,
> > > Aaron
> > >
> > > Quoting Matthew Johnson :
> > >
> > > > We had a lot of trouble with the Java version of Aria as well.
> > > Which
> > > > version of Aria are you using?
> > > >
> > > > We have it working under Aria2.5.1 and jdk1.5.0_09. If you
> think
> > > > that would
> > > > work I can send it to you.
> > > >
> > > > Matt
> > > >
> > > > -----Original Message-----
> > > > From: aria-users-bounces at lists.mobilerobots.com
> > > > [mailto:aria-users-bounces at lists.mobilerobots.com] On Behalf Of
> > > > aaron123 at ksu.edu
> > > > Sent: Tuesday, June 24, 2008 12:51 PM
> > > > To: aria-users at mobilerobots.com
> > > > Subject: [Aria-users] Java 1.6 wrapper update
> > > >
> > > > Has anyone created a recent JNI wrapper for Aria? Our project
> is
> > > > using
> > > > Java 1.6 but using the current JNI wrapper causes quirky (and
> > > > frustrating) errors. I've attempted recompiling the project on
> > > > Windows
> > > > using Visual Studio and SWIG but the resulting files don't work
> > > > properly. As I am not an expert with SWIG, this is making
> > progress
> > > > on
> > > > the project difficult. Much of the code written uses Java 1.6
> > > > constructs so reverting back to Java 1.4.2 isn't an option.
> > > >
> > > > The main problem we are having is when dealing with the ArSick
> > > data.
> > > > We
> > > > require the raw data from the laser and the only way to obtain
> it
> > > is
> > > > using the getRawReadingsAsVector method. From there, we using
> > the
> > > > ArSensorReadingVector.get method to obtain an ArSensorReading
> > from
> > > > which we get the range value. This works for about 30 seconds
> > then
> > > > for
> > > > some reason a JVM call is made...
> > > > AriaJavaJNI.delete_ArSerialConnection
> > > > .. which kills the serial connection and crashes the app. I
> was
> > > able
> > > > to find this problem using the -verbose:jni -Xcheck:jni VM
> > options
> > > > when
> > > > running the program. I have confirmed that when compiling and
> > > > running
> > > > the laser code in isolation using Java 1.4.2 it runs fine.
> > > >
> > > > Assistance would be much appreciated.
> > > >
> > > >
> > > > Thanks,
> > > > Aaron
> > > >
> > > > _______________________________________________
> > > > Aria-users mailing list
> > > > Aria-users at lists.mobilerobots.com
> > > > http://lists.mobilerobots.com/mailman/listinfo/aria-users
> > > >
> > > > To unsubscribe visit the above webpage or send an e-mail to:
> > > >
> > > > aria-users-leave at lists.mobilerobots.com
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> >
> >
> >
> >
>
>
>
>
>
From Jason.Chen at anu.edu.au Fri Jun 27 01:35:22 2008
From: Jason.Chen at anu.edu.au (Jason Chen)
Date: Fri, 27 Jun 2008 15:35:22 +1000
Subject: [Aria-users] laser integration pcb
In-Reply-To: <485658D1.407@mobilerobots.com>
References: <284418.49305.qm@web36103.mail.mud.yahoo.com>
<200806161818.31575.Jason.Chen@anu.edu.au>
<485658D1.407@mobilerobots.com>
Message-ID: <200806271535.22356.Jason.Chen@anu.edu.au>
Hi Reed,
I have a question about the "laser integration" board that came with my DX3.
It has a P/N ELC1134 and Rev E.
I'm wanting to run another laser (ie 2 x LMS200s on the robot). This pcb from
what I can tell is simply a DC to DC converter to get 24V from 12V. So the
first question regards power: Can the DC to DC converter supply enough
current for 2 lasers?
The other question regards communication. The LMS200 is connected to the pcb
by 3 wires, ie RxD, TxD and Gnd - this makes sense. The pcb is then
connected back to the com3 port on the VSBC-8 onboard computer with 4 wires.
The fourth wire by my account is the DTR line, which I'm gathering turns on
the DC to DC converter using the BJT's on the pcb when the serial port
becomes ready. Is this correct?
Is there any documentation around for this board?
I looked around for some, but couldn't find any
Cheers
Jason
From Jason.Chen at anu.edu.au Fri Jun 27 01:35:22 2008
From: Jason.Chen at anu.edu.au (Jason Chen)
Date: Fri, 27 Jun 2008 15:35:22 +1000
Subject: [Aria-users] laser integration pcb
In-Reply-To: <485658D1.407@mobilerobots.com>
References: <284418.49305.qm@web36103.mail.mud.yahoo.com>
<200806161818.31575.Jason.Chen@anu.edu.au>
<485658D1.407@mobilerobots.com>
Message-ID: <200806271535.22356.Jason.Chen@anu.edu.au>
Hi Reed,
I have a question about the "laser integration" board that came with my DX3.
It has a P/N ELC1134 and Rev E.
I'm wanting to run another laser (ie 2 x LMS200s on the robot). This pcb from
what I can tell is simply a DC to DC converter to get 24V from 12V. So the
first question regards power: Can the DC to DC converter supply enough
current for 2 lasers?
The other question regards communication. The LMS200 is connected to the pcb
by 3 wires, ie RxD, TxD and Gnd - this makes sense. The pcb is then
connected back to the com3 port on the VSBC-8 onboard computer with 4 wires.
The fourth wire by my account is the DTR line, which I'm gathering turns on
the DC to DC converter using the BJT's on the pcb when the serial port
becomes ready. Is this correct?
Is there any documentation around for this board?
I looked around for some, but couldn't find any
Cheers
Jason
From aaron123 at ksu.edu Fri Jun 27 10:11:50 2008
From: aaron123 at ksu.edu (aaron123 at ksu.edu)
Date: Fri, 27 Jun 2008 09:11:50 -0500
Subject: [Aria-users] ACTS using uncompressed video data
In-Reply-To: <1213810775.4859485800593@webmail.ksu.edu>
References: <1213217300.48503a1475f69@webmail.ksu.edu>
<4858273F.1060002@mobilerobots.com>
<1213810775.4859485800593@webmail.ksu.edu>
Message-ID: <1214575910.4864f5266be72@webmail.ksu.edu>
Hi Reed. We are using the Sensoray 311 framegrabber. I've captured
images taken directly from the framegrabber and those used by ACTS
and
I've attached them for your reference. As you'll see, most of the
image
is fine but the object that is being detected contains a checkerboard
grid of compression artifacts(I had to remove them because the email was
being blocked with the attachments). Typically this isn't a huge deal
but
we're working with these things outdoors and at a distance of about
56"
the artifacts interfere with the blob detection. I thought it would
be
nice to use the raw framegrabber image in the ACTS software to
enhance
the detection distance but it appears that ACTS compresses the data
first then does the blob detection on that image (I understand the
reasons for the compression). I have alternative methods in mind if
this isn't feasible but I thought I'd ask an expert first before
discounting the idea.
Thanks for your assistance,
Aaron
>
> Quoting Reed Hedges :
>
> >
> > Hi Aaron,
> >
> > What do you mean by uncompressed data from the camera? What kind
> of
> > framegrabber are you using?
> >
> >
> >
> > There may be some framegrabber-specific options available... check
> > the Video
> > Parameters dialog, or run ACTS on a command line with the -P option
> > to see a
> > list of them. Then you can run it with one of those options set on
> a
> > command
> > line by including the -s option. See Table 1 in chapter 2 of the
> ACTS
> > manual.
> >
> > You can have ACTS work on a single image you provide (Load Image in
> > the menu),
> > but not a stream of images.
> >
> > If simply selecting a slower rate helps, you can change that too...
> > though I
> > think that just changes the rate at which ACTS acquires the images
> > from the
> > framegrabber.
> >
> >
> > Reed
> >
> >
> > aaron123 at ksu.edu wrote:
> > > Hello everyone. I've read the ACTS manual and there doesn't
> appear
> > to
> > > be a way to run ACTS using uncompressed data coming from the
> camera
> > but
> > > I would really like to do this even if it means using a slower
> > frame
> > > rate. The other alternative is to use ACTS to process images
> > that I
> > > pass to it although I see no documentation on this type of
> feature
> > > which leads me to believe that it isn't possible. Does anyone
> have
> > any
> > > ideas on a good approach/option?
> >
> >
> > _______________________________________________
> > Aria-users mailing list
> > Aria-users at lists.mobilerobots.com
> > http://lists.mobilerobots.com/mailman/listinfo/aria-users
> >
> > To unsubscribe visit the above webpage or send an e-mail to:
> >
> > aria-users-leave at lists.mobilerobots.com
> >
> >
>
>
>
From reed at mobilerobots.com Fri Jun 27 10:37:40 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 27 Jun 2008 10:37:40 -0400
Subject: [Aria-users] ACTS using uncompressed video data
In-Reply-To: <1214575910.4864f5266be72@webmail.ksu.edu>
References: <1213217300.48503a1475f69@webmail.ksu.edu>
<4858273F.1060002@mobilerobots.com>
<1213810775.4859485800593@webmail.ksu.edu>
<1214575910.4864f5266be72@webmail.ksu.edu>
Message-ID: <4864FB34.2060605@mobilerobots.com>
Hi Aaron,
Can you tell if it's doing the compression when it reads the image? (It could
just be the way that ACTS displays the image in the GUI.)
I'll look at the ACTS source code at some point, not sure when I'll be able to
get to it though.
Reed
aaron123 at ksu.edu wrote:
> Hi Reed. We are using the Sensoray 311 framegrabber. I've captured
> images taken directly from the framegrabber and those used by ACTS
> and
> I've attached them for your reference. As you'll see, most of the
> image
> is fine but the object that is being detected contains a checkerboard
> grid of compression artifacts(I had to remove them because the email was
> being blocked with the attachments). Typically this isn't a huge deal
> but
> we're working with these things outdoors and at a distance of about
> 56"
> the artifacts interfere with the blob detection. I thought it would
> be
> nice to use the raw framegrabber image in the ACTS software to
> enhance
> the detection distance but it appears that ACTS compresses the data
> first then does the blob detection on that image (I understand the
> reasons for the compression). I have alternative methods in mind if
> this isn't feasible but I thought I'd ask an expert first before
> discounting the idea.
>
> Thanks for your assistance,
> Aaron
>> Quoting Reed Hedges :
>>
>>> Hi Aaron,
>>>
>>> What do you mean by uncompressed data from the camera? What kind
>> of
>>> framegrabber are you using?
>>>
>>>
>>>
>>> There may be some framegrabber-specific options available... check
>>> the Video
>>> Parameters dialog, or run ACTS on a command line with the -P option
>>> to see a
>>> list of them. Then you can run it with one of those options set on
>> a
>>> command
>>> line by including the -s option. See Table 1 in chapter 2 of the
>> ACTS
>>> manual.
>>>
>>> You can have ACTS work on a single image you provide (Load Image in
>>> the menu),
>>> but not a stream of images.
>>>
>>> If simply selecting a slower rate helps, you can change that too...
>>> though I
>>> think that just changes the rate at which ACTS acquires the images
>>> from the
>>> framegrabber.
>>>
>>>
>>> Reed
>>>
>>>
>>> aaron123 at ksu.edu wrote:
>>>> Hello everyone. I've read the ACTS manual and there doesn't
>> appear
>>> to
>>>> be a way to run ACTS using uncompressed data coming from the
>> camera
>>> but
>>>> I would really like to do this even if it means using a slower
>>> frame
>>>> rate. The other alternative is to use ACTS to process images
>>> that I
>>>> pass to it although I see no documentation on this type of
>> feature
>>>> which leads me to believe that it isn't possible. Does anyone
>> have
>>> any
>>>> ideas on a good approach/option?
>>>
From reed at mobilerobots.com Fri Jun 27 10:41:08 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 27 Jun 2008 10:41:08 -0400
Subject: [Aria-users] Java 1.6 wrapper update - progress update
In-Reply-To: <1214505487.4863e20f98459@webmail.ksu.edu>
References: <1214329874.48613412dc0f8@webmail.ksu.edu> <000601c8d62b$d87a7160$1a00640a@mjdell> <1214335730.48614af2d582e@webmail.ksu.edu> <000301c8d641$877a32d0$3b00640a@mjdell> <1214343638.486169d61e440@webmail.ksu.edu> <000001c8d66c$5f481270$6401a8c0@mjdell> <1214419912.486293c88603b@webmail.ksu.edu> <000601c8d6f8$e5c6e5b0$1a00640a@mjdell>
<1214505487.4863e20f98459@webmail.ksu.edu>
Message-ID: <4864FC04.9030408@mobilerobots.com>
aaron123 at ksu.edu wrote:
> I have found
> that using these VM options at run time "-Xcheck:jni -verbose:gc
> -verbose:jni ", I have been able to isolate where and what errant JNI
> calls are made and can add robustness code which eliminates the
> crashes. I've even been able to run different versions of the Swig
> wrapper with newer versions of Java in this way.
If you have a log of this information handy, could I take a look at it? There
may be some things we can do in the SWIG wrapper specification to avoid them
(not sure what yet though).
Reed
From aaron123 at ksu.edu Fri Jun 27 10:50:36 2008
From: aaron123 at ksu.edu (Aaron Chleborad)
Date: Fri, 27 Jun 2008 09:50:36 -0500
Subject: [Aria-users] Aria Java Wrapper Tip
In-Reply-To: <1214505487.4863e20f98459@webmail.ksu.edu>
References: <1214329874.48613412dc0f8@webmail.ksu.edu>
<000601c8d62b$d87a7160$1a00640a@mjdell>
<1214335730.48614af2d582e@webmail.ksu.edu>
<000301c8d641$877a32d0$3b00640a@mjdell>
<1214343638.486169d61e440@webmail.ksu.edu>
<000001c8d66c$5f481270$6401a8c0@mjdell>
<1214419912.486293c88603b@webmail.ksu.edu>
<000601c8d6f8$e5c6e5b0$1a00640a@mjdell>
<1214505487.4863e20f98459@webmail.ksu.edu>
Message-ID: <1214578236.4864fe3cee655@webmail.ksu.edu>
For those frustrated with "pure virtual function call" errors using the
Aria jni wrapper I've found the following to be helpful (this is on a
Windows XP system and using eclipse but it may work on other systems).
Use the following Java VM options at run time "-Xcheck:jni -verbose:gc
-verbose:jni " and look for strange jni calls that are crashing the
program. I've found that Java's garbage collection mechanism seems to
delete jni objects improperly occasionally. There are a few methods
that can be used to solve the problem.
1. Rebuild the jni wrapper and hope that it works (this is not fun).
2. Based on the vm options above, determine which objects are being
garbage collected improperly and declare them as static in your class.
This should prevent the GC from doing nasty things to it.
3. Add code to check to see if your object is null and if so create a
new object before it is used again.
I realize the ideal method is to rebuild the jni wrapper but this can be
time consuming and difficult and not everyone has the luxury of doing
this so in a pinch methods 2 and 3 can alleviate the "pure virtual
function call" errors.
Hope this info helps someone. If it does, smack me an email.
--
Aaron Chleborad
Kansas State University
Department of Computing and Information Sciences
Research Assistant
http://macr.cis.ksu.edu/
a a r o n 1 2 3 (at) k s u (tiny dot) e d u
From reed at mobilerobots.com Fri Jun 27 11:14:21 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 27 Jun 2008 11:14:21 -0400
Subject: [Aria-users] laser integration pcb
Message-ID: <486503CD.6060702@mobilerobots.com>
-------- Original Message --------
Subject: [Support] Re: [Aria-users] laser integration pcb
Date: Fri, 27 Jun 2008 10:44:27 -0400
From: Mobile Robots/ActivMedia Support
Hello Jason,
The DC-DC converter for the laser is only rated for 25 W output power,
so it will not power more than one laser (the lasers typically use a bit
less than 1 A or ~20 W). You would need another DC-DC to run another
laser.
You are correct about the function of the fourth conductor from the
serial port. The laser power is automatically switched on when the
serial port is opened. We do this because if the laser is not powered
off between connections, it can occasionally get into a state where it
will not connect.
Best,
-Zeb Dahl
On Fri, 2008-06-27 at 15:35 +1000, Jason Chen wrote:
> Hi Reed,
> I have a question about the "laser integration" board that came with my DX3.
> It has a P/N ELC1134 and Rev E.
>
> I'm wanting to run another laser (ie 2 x LMS200s on the robot). This pcb from
> what I can tell is simply a DC to DC converter to get 24V from 12V. So the
> first question regards power: Can the DC to DC converter supply enough
> current for 2 lasers?
>
> The other question regards communication. The LMS200 is connected to the pcb
> by 3 wires, ie RxD, TxD and Gnd - this makes sense. The pcb is then
> connected back to the com3 port on the VSBC-8 onboard computer with 4 wires.
> The fourth wire by my account is the DTR line, which I'm gathering turns on
> the DC to DC converter using the BJT's on the pcb when the serial port
> becomes ready. Is this correct?
>
> Is there any documentation around for this board?
> I looked around for some, but couldn't find any
>
> Cheers
> Jason
> aria-users-leave at lists.mobilerobots.com
>
From reed at mobilerobots.com Fri Jun 27 11:23:41 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 27 Jun 2008 11:23:41 -0400
Subject: [Aria-users] video server
In-Reply-To: <1672c7d80806191000o71d54dsa881eaa8a6af3772@mail.gmail.com>
References: <1672c7d80804101152k3856e1e4u5ae50252291a252d@mail.gmail.com> <4803CDA4.1010101@mobilerobots.com> <1672c7d80806121156h69f4b29at760364f4846315d7@mail.gmail.com> <48528E1D.7050902@mobilerobots.com> <1672c7d80806170140p1ec7ec52n86e95521d185363d@mail.gmail.com> <485824E8.6050004@mobilerobots.com>
<1672c7d80806191000o71d54dsa881eaa8a6af3772@mail.gmail.com>
Message-ID: <486505FD.7080403@mobilerobots.com>
Hi Alberto,
What SAV server does is compress the image to a JPEG and send that using
ArNetworking. You can use libjpeg to do that on Linux (it's a standard library,
you can install it from RedHat or use apt-get on Debian).
To transfer the images using ArNetworking, on the server set up a packet handler
to respond to "getPictureCam1" requests. In each reply packet put a pair of
16-bit integers containing image width and height in pixels using uByte2ToBuf(),
then the JPEG data using dataToBuf(). The ArNetworking example
getVideoExample.cpp requests images using getPictureCam1 and this packet format.
MobileEyes does too, so you can use that as your client if you just want to
see the image (you could also then use ArConfig to store any parameters and use
ArServerHandlerConfig to let MobileEyes modify the values in ArConfig too.)
If I have time later I'll write an example which does this. But if you look at
the other ArNetworking example servers which respond to packets you should be
able to do the same thing, but put the image width, height and JPEG data in it
instead.
Reed
Alberto Corrales Garcia wrote:
> Hello and thank you for your help,
> I'm learning about Aria and ArNetworking but I'm a little confusing,
> because some subjects are different to the traditional C++ programming.
>
> I would be very pleased if you could send me code for a server video or
> tell me how I can respond to the video requests.
>
> Regards
>
>
> 2008/6/17 Reed Hedges >:
>
> Alberto Corrales Garcia wrote:
> > yes, I think if the source and the receiver would be synchronize
> sendig
> > frames to see the video correctly. I saw the class
> > ArHybridForwarderVideo, which has some methods to manage that
> subject,
> > but I don't know how it work and I didn't see any example. Could you
> > tell me about that class?
>
> Hi Alberto,
>
> The client will receive each packet from the server (source) in turn, at
> whatever rate it requests. For example if you call:
>
> client.addHandler("getPictureCam1", &handlerFunctor);
> client.request("getPictureCam1", 30);
>
> Then the client will send the getPictureCam1 request to the server
> 30 times a
> second. And the client's callback function will be called serially
> for each
> response packet is received (one thread).
>
> When the server receives a request it may then read an image from the
> framegrabber, or it may have an image stored from an asynchronous
> thread... I
> don't remember how exactly it works right now but will look at some
> of the code
> to see if I can determine that.
>
> However, in either case the server won't get "backed up"-- there is
> no queue of
> images waiting to be sent, it's always the most recent one it has.
> (How your
> client program handles the images is up to you though.) It is
> possible to
> "miss" a potentially unique image, but you can minimize this by
> using a request
> rate faster than the framegrabber's own rate of generating images
> (you will get
> some duplicate images in that case).
>
> So the client is synchronized with the server by the
> request-response mechanism,
> but this is may be asynchronous in a sense from how the framegrabber
> or other
> source is providing images or compared to the ideal sequence of
> unique images
> you could be getting.
>
> ArHybridForwarderVideo is a class used to forward ArNetworking video
> requests
> and data obtained from another ArNetworkingServer. For example if
> you start
> ACTS or SAVServer, it provides video images on port 7070 using
> ArNetworking. If
> you then start ARNL's guiServer, it contains an
> ArHybridForwarderVideo object,
> which grabs the images from ACTS or SAVServer, and turns around and
> provides
> them on the same port as the rest of its services (robot info, path
> planning,
> map, config, etc.). Then you can connect with MobileEyes and get
> all the
> services, including video images, from the same server and port.
>
> Does this make sense? Let me know if there are any details you
> need me to find
> out.
>
> Reed
>
From lafary at mobilerobots.com Fri Jun 27 16:28:44 2008
From: lafary at mobilerobots.com (Matt LaFary)
Date: Fri, 27 Jun 2008 16:28:44 -0400
Subject: [Aria-users] ACTS using uncompressed video data
In-Reply-To: <4864FB34.2060605@mobilerobots.com>
References: <1213217300.48503a1475f69@webmail.ksu.edu> <4858273F.1060002@mobilerobots.com> <1213810775.4859485800593@webmail.ksu.edu> <1214575910.4864f5266be72@webmail.ksu.edu>
<4864FB34.2060605@mobilerobots.com>
Message-ID: <48654D7C.50102@mobilerobots.com>
I'm 99% certain ACTS operates on an uncompressed image when operating in
normal read the framegrabber mode and its never compressed before
analysis. It might compress it for display, and it certainly is
compressed if you're viewing it in MobileEyes or SAV viewer or what not.
You could send Reed and I your attachments for this directly though and
we'll take a look at it. It'd be most helpful if you had some
screenshots of what acts was reporting detected too.
Matt LaFary
MobileRobots Inc
Reed Hedges wrote:
> Hi Aaron,
>
> Can you tell if it's doing the compression when it reads the image? (It could
> just be the way that ACTS displays the image in the GUI.)
>
> I'll look at the ACTS source code at some point, not sure when I'll be able to
> get to it though.
>
> Reed
>
>
>
> aaron123 at ksu.edu wrote:
>> Hi Reed. We are using the Sensoray 311 framegrabber. I've captured
>> images taken directly from the framegrabber and those used by ACTS
>> and
>> I've attached them for your reference. As you'll see, most of the
>> image
>> is fine but the object that is being detected contains a checkerboard
>> grid of compression artifacts(I had to remove them because the email was
>> being blocked with the attachments). Typically this isn't a huge deal
>> but
>> we're working with these things outdoors and at a distance of about
>> 56"
>> the artifacts interfere with the blob detection. I thought it would
>> be
>> nice to use the raw framegrabber image in the ACTS software to
>> enhance
>> the detection distance but it appears that ACTS compresses the data
>> first then does the blob detection on that image (I understand the
>> reasons for the compression). I have alternative methods in mind if
>> this isn't feasible but I thought I'd ask an expert first before
>> discounting the idea.
>>
>> Thanks for your assistance,
>> Aaron
>>> Quoting Reed Hedges :
>>>
>>>> Hi Aaron,
>>>>
>>>> What do you mean by uncompressed data from the camera? What kind
>>> of
>>>> framegrabber are you using?
>>>>
>>>>
>>>>
>>>> There may be some framegrabber-specific options available... check
>>>> the Video
>>>> Parameters dialog, or run ACTS on a command line with the -P option
>>>> to see a
>>>> list of them. Then you can run it with one of those options set on
>>> a
>>>> command
>>>> line by including the -s option. See Table 1 in chapter 2 of the
>>> ACTS
>>>> manual.
>>>>
>>>> You can have ACTS work on a single image you provide (Load Image in
>>>> the menu),
>>>> but not a stream of images.
>>>>
>>>> If simply selecting a slower rate helps, you can change that too...
>>>> though I
>>>> think that just changes the rate at which ACTS acquires the images
>>>> from the
>>>> framegrabber.
>>>>
>>>>
>>>> Reed
>>>>
>>>>
>>>> aaron123 at ksu.edu wrote:
>>>>> Hello everyone. I've read the ACTS manual and there doesn't
>>> appear
>>>> to
>>>>> be a way to run ACTS using uncompressed data coming from the
>>> camera
>>>> but
>>>>> I would really like to do this even if it means using a slower
>>>> frame
>>>>> rate. The other alternative is to use ACTS to process images
>>>> that I
>>>>> pass to it although I see no documentation on this type of
>>> feature
>>>>> which leads me to believe that it isn't possible. Does anyone
>>> have
>>>> any
>>>>> ideas on a good approach/option?
>
>
> _______________________________________________
> Aria-users mailing list
> Aria-users at lists.mobilerobots.com
> http://lists.mobilerobots.com/mailman/listinfo/aria-users
>
> To unsubscribe visit the above webpage or send an e-mail to:
>
> aria-users-leave at lists.mobilerobots.com
>
From reed at mobilerobots.com Fri Jun 27 16:55:52 2008
From: reed at mobilerobots.com (Reed Hedges)
Date: Fri, 27 Jun 2008 16:55:52 -0400
Subject: [Aria-users] ACTS using uncompressed video data
In-Reply-To: <1214581508.48650b04722d0@webmail.ksu.edu>
References: <1213217300.48503a1475f69@webmail.ksu.edu>
<4858273F.1060002@mobilerobots.com>
<1213810775.4859485800593@webmail.ksu.edu>
<1214575910.4864f5266be72@webmail.ksu.edu>
<4864FB34.2060605@mobilerobots.com>
<1214581508.48650b04722d0@webmail.ksu.edu>
Message-ID: <486553D8.2000707@mobilerobots.com>
Aaron Chleborad wrote:
> Essentially, I unable to determine if acts processes the original image
> or the modified version before sending it to the gui.
I'm pretty sure that the artifacts are from the GUI display only (it's scaling
the image down, and also X and the GUI toolkit may be doing something to the
image to display it efficiently or whatever) and that the color finding is done
on the original image.
In fact you can even turn the GUI off I think with the -t command line option.
When I have time I'll try to verify this though, just to be sure.
Reed
--
Reed Hedges
Developer Software Support
MobileRobots Inc.
Find downloads, documentation, FAQ, mailing lists and more at: