From jcallen1 at ithaca.edu  Thu Jun  8 11:33:57 2006
From: jcallen1 at ithaca.edu (James Callender)
Date: Thu Jun  8 11:33:58 2006
Subject: [Pyro-users] ARIA support
Message-ID: <50749.147.129.100.218.1149780837.squirrel@webmail.ithaca.edu>

I've looked through some of the mailing list archives, and it seems like
there used to be ARIA support for pyrobot.  I'm currently doing an
internship and we are required to use the ARIA software installed on our
Pioneer3 robot.  Is the code still available for the ARIA interface?

- james

From fox at macalester.edu  Thu Jun  8 17:14:02 2006
From: fox at macalester.edu (Susan Fox)
Date: Thu Jun  8 17:12:50 2006
Subject: [Pyro-users] Installing Pyro...
Message-ID: <4488931A.1060805@macalester.edu>

An HTML attachment was scrubbed...
URL: http://emergent.brynmawr.edu/pipermail/pyro-users/attachments/20060608/8d79e849/attachment.htm
From dblank at brynmawr.edu  Thu Jun  8 17:35:12 2006
From: dblank at brynmawr.edu (dblank@brynmawr.edu)
Date: Thu Jun  8 17:35:14 2006
Subject: [Pyro-users] Installing Pyro...
In-Reply-To: <4488931A.1060805@macalester.edu>
References: <4488931A.1060805@macalester.edu>
Message-ID: <41073.68.34.179.248.1149802512.squirrel@webmail.brynmawr.edu>

Hi Susan,

What distribution of Linux are you using? python-imaging should be a
standard package available from the "extras" repository. When you type:

yum list | grep python-imaging

what do you get?

-Doug

> Hi,<br>
> <br>
> The python-imaging package doesn't seem to be available where the
> installation information suggests that it should be.&nbsp; Has something
> changed that makes the installation instructions out of date?<br>
> <br>
> I was trying to use yum, but it couldn't find python-imaging, and so
> failed to install pyrobot itself.<br>
> <br>
> Thanks for the help,<br>
> <br>
> Susan <br>
> </font></font>
> <pre class="moz-signature" cols="72">--
> Susan Fox                                 Associate Professor
> Mathematics/Computer Science              Macalester College
> Office: (651) 696-6553                    Fax: (651) 696-6518
> <a class="moz-txt-link-abbreviated"
> href="mailto:fox@macalester.edu">fox@macalester.edu</a>
>     <a class="moz-txt-link-freetext"
> href="http://www.macalester.edu/~fox">http://www.macalester.edu/~fox</a>
>
> </pre>
> </body>
> </html>
> _______________________________________________
> Pyro-users mailing list
> Pyro-users@emergent.brynmawr.edu
> http://emergent.brynmawr.edu/mailman/listinfo/pyro-users
>


From dblank at brynmawr.edu  Mon Jun 12 10:04:49 2006
From: dblank at brynmawr.edu (Douglas S. Blank)
Date: Mon Jun 12 10:05:25 2006
Subject: [Pyro-users] ARIA support
In-Reply-To: <50749.147.129.100.218.1149780837.squirrel@webmail.ithaca.edu>
References: <50749.147.129.100.218.1149780837.squirrel@webmail.ithaca.edu>
Message-ID: <1150121089.29896.36.camel@mightymouse.brynmawr.edu>

James,

Because we can communicate with the Pioneers through Player, we don't
need the Aria interface anymore. It was really just using Aria as a
pass-through.

Is there something that Aria provided that Player does not? The current
version of Pyro (4.x) uses Player 1.6, and Pyro 5.0 will use Player 2.0.
You should be able to just install Player, and Pyro will talk to your
robot through that. We use Stage, which MobleSim is based on, and Stage
is Player-compatible.

If you really want to reintegrate Aria with Pyro, we can give you some
pointers. It wouldn't be that much work, but it is redundant with the
Player interface. (You need to use an extra cable if you use Player to
for the serial port, so an Aria interface would save you from getting
the cable, but maintaining two interfaces is, well, twice as much work
as maintaining one.) 

Hope that helps,

-Doug

On Thu, 2006-06-08 at 11:33 -0400, James Callender wrote:
> I've looked through some of the mailing list archives, and it seems like
> there used to be ARIA support for pyrobot.  I'm currently doing an
> internship and we are required to use the ARIA software installed on our
> Pioneer3 robot.  Is the code still available for the ARIA interface?
> 
> - james
> 
> _______________________________________________
> Pyro-users mailing list
> Pyro-users@emergent.brynmawr.edu
> http://emergent.brynmawr.edu/mailman/listinfo/pyro-users
> 
-- 
Douglas S. Blank       Computer Science
Assistant Professor    Bryn Mawr College
(610)526-6501          http://cs.brynmawr.edu/~dblank


From fox at macalester.edu  Thu Jun 15 14:10:18 2006
From: fox at macalester.edu (Susan Fox)
Date: Thu Jun 15 14:08:53 2006
Subject: [Pyro-users] example doesn't work...
Message-ID: <4491A28A.3090807@macalester.edu>

An HTML attachment was scrubbed...
URL: http://emergent.brynmawr.edu/pipermail/pyro-users/attachments/20060615/bc6d837e/attachment.htm
From dblank at brynmawr.edu  Thu Jun 15 17:29:30 2006
From: dblank at brynmawr.edu (Douglas S. Blank)
Date: Thu Jun 15 17:29:36 2006
Subject: [Pyro-users] example doesn't work...
In-Reply-To: <4491A28A.3090807@macalester.edu>
References: <4491A28A.3090807@macalester.edu>
Message-ID: <1150406970.2703.194.camel@mightymouse.brynmawr.edu>

Susan,

Yes, you are right: the new version doesn't need the select() function
anymore. Those "import ... select"  and "import ... avg" lines should
not have been there, and have now been removed from the wiki pages. The
code had been updated, but the imports hadn't been removed. Thanks for
catching those!

To convert your old code to the new Pythonic interfaces and syntax, see:

http://pyrorobotics.org/?page=PythonicPyroIntro

The older version required that you jump through too many hoops:

>>> from pyrobot.brain import select, avg 
>>> select(min, "value", self.get("/robot/range/all/pos,thr,value"))

The newer interface is very Python-oriented:

>>> min([(sensor.distance(), sensor.angle()) for sensor in robot.range])

This new version uses "list comprehension" and now makes each
sensor-reading an object that has properties and methods (such as
distance() and angle() in the above). "min" is now just the Python
minimum function applied to tuples of (distance,angle) for each reading.

Hope that helps,

-Doug

On Thu, 2006-06-15 at 13:10 -0500, Susan Fox wrote:
> 
> Hi, there,
> 
> I'm working with the behavior-based control stuff, and am having a
> minor problem, now that I've upgraded to the current version of Pyro
> (pyrobot).  My code used the 'select' command from pyro.brain, and it
> doesn't appear to be in the new version.  However, Exercise 2 in the
> Behavior-Based Control curriculum module indicates that "from
> pyrobot.brain import select" should work.  On my system, it doesn't:
> that example doesn't load and says that there is no such thing as
> 'select'.  Any ideas what I might be missing?
> 
> Thanks,
> 
> Susan
> -- 
> Susan Fox                                 Associate Professor
> Mathematics/Computer Science              Macalester College
> Office: (651) 696-6553                    Fax: (651) 696-6518
> fox@macalester.edu                        http://www.macalester.edu/~fox
> 
> _______________________________________________
> Pyro-users mailing list
> Pyro-users@emergent.brynmawr.edu
> http://emergent.brynmawr.edu/mailman/listinfo/pyro-users
-- 
Douglas S. Blank       Computer Science
Assistant Professor    Bryn Mawr College
(610)526-6501          http://cs.brynmawr.edu/~dblank


From fox at macalester.edu  Fri Jun 16 16:09:26 2006
From: fox at macalester.edu (Susan Fox)
Date: Fri Jun 16 16:07:57 2006
Subject: [Pyro-users] Coordinate issues with Player and Stage (and Pyro)
Message-ID: <44930FF6.4040201@macalester.edu>

An HTML attachment was scrubbed...
URL: http://emergent.brynmawr.edu/pipermail/pyro-users/attachments/20060616/14227b05/attachment.htm
From dblank at brynmawr.edu  Mon Jun 19 01:26:13 2006
From: dblank at brynmawr.edu (dblank@brynmawr.edu)
Date: Mon Jun 19 01:26:16 2006
Subject: [Pyro-users] Coordinate issues with Player and Stage (and Pyro)
In-Reply-To: <44930FF6.4040201@macalester.edu>
References: <44930FF6.4040201@macalester.edu>
Message-ID: <48735.68.34.179.248.1150694773.squirrel@webmail.brynmawr.edu>

Susan,

You're right again: it appears that the old Player robot code did not
reset the x,y,a correctly. That will be corrected in the new Player code.
I would suggest to try out the new Pyrobot Simulator. (I just noticed that
there was a little error in that localize code, but was easily fixed just
now.)

So, to try out your mapping code, I would update your Pyrobot install to
include the new code, and try the Pyrobot Simulator/PyrobotRobot60000.
Here is the easiest way to update your install (assumes a FC5 yum install
in the standard place):

$ su -
# cd /usr/lib/python2.4/site-packages/
# mv pyrobot pyrobot.bak
# cvs -d :pserver:anonymous@compscitest.brynmawr.edu:/cvs login
Password: <JUST HIT ENTER>
# cvs -d :pserver:anonymous@compscitest.brynmawr.edu:/cvs co pyrobot
...
# cp pyrobot.bak/Makefile.cfg pyrobot/
# cd pyrobot
# make

This also fixes a memory leak that we had in Pyrobot Simulator.

We are getting close to having the new Player/Stage working, too.

Thangs for the bug reports; hope this gets you going!

-Doug

> I'm using the new version of Pyro, and have run into two semi-related
> issues having to do with coordinate systems that, so far, have stymied
> me, at least in terms of finding *easy* solutions.<br>
> <br>
> I have a navigation system that I developed under the old Pyro, using a
> particular coordinate map for the robot's space.&nbsp; Now, the system
> wants
> to set the origin to wherever the robot is when it starts up, and to
> report its position as an offset to that point.&nbsp; However, the robot
> starts up in many different positions, and I need to be able to
> maintain its position in a global coordinate system, and to allow a
> user to specify the robot's position.&nbsp; I've tried using
> robot.localize(), but that is overridden by the Player.py update()
> method.&nbsp; What is my easiest solution: 1) keeping track of the global
> coordinate within the robot's brain and updating it manually or, 2)
> making a subclass of the PlayerRobot class and changing the way the
> robot's coordinates are set.&nbsp; Has anyone else dealt with this in an
> elegant way?<br>
> <br>
> The second issue has to do with Stage's imposed coordinate system.&nbsp;
> It
> doesn't map at all onto the coordinate system&nbsp; that my system has
> been
> using.&nbsp; Is it possible to alter Stage's coordinate system, or is my
> best bet to ignore Stage's system altogether?&nbsp; <br>
> <br>
> Thanks for any good advice you have.&nbsp; :)<br>
> <br>
> Susan Fox<br>
> </font></font>
> <pre class="moz-signature" cols="72">--
> Susan Fox                                 Associate Professor
> Mathematics/Computer Science              Macalester College
> Office: (651) 696-6553                    Fax: (651) 696-6518
> <a class="moz-txt-link-abbreviated"
> href="mailto:fox@macalester.edu">fox@macalester.edu</a>
>     <a class="moz-txt-link-freetext"
> href="http://www.macalester.edu/~fox">http://www.macalester.edu/~fox</a>
>
> </pre>
> </body>
> </html>
> _______________________________________________
> Pyro-users mailing list
> Pyro-users@emergent.brynmawr.edu
> http://emergent.brynmawr.edu/mailman/listinfo/pyro-users
>


From braught at dickinson.edu  Tue Jun 27 16:09:32 2006
From: braught at dickinson.edu (Grant Braught)
Date: Tue Jun 27 16:09:41 2006
Subject: [Pyro-users] Custom FakeCamera Images
Message-ID: <c1a311c4f0115819198dab334dedaa0d@dickinson.edu>

I'll be using Pyro in my AI course in the Fall and I've just begun work 
on some homework exercises for the Vision unit.  What I want is for the 
students to be able to write a brain that creates a FakeCamera and 
points it to a directory containing a bunch of ppm image files that 
would then be displayed and processed in much the same way the 
FakeCamrea and the TutorialCamera work.  I saw the page on "Simulated 
Vision Using FakeCamera" but it seems to be missing some information.  
The second paragraph refers to two examples that would seem to be 
exactly what I want but the examples don't seem to be there.

Any suggestions for or examples of how to create a FakeCamera using a 
custom set of images within a brain would be greatly appreciated.

Thanks,
Grant



From dblank at brynmawr.edu  Tue Jun 27 17:23:26 2006
From: dblank at brynmawr.edu (Douglas S. Blank)
Date: Tue Jun 27 17:23:23 2006
Subject: [Pyro-users] Custom FakeCamera Images
In-Reply-To: <c1a311c4f0115819198dab334dedaa0d@dickinson.edu>
References: <c1a311c4f0115819198dab334dedaa0d@dickinson.edu>
Message-ID: <1151443406.2777.34.camel@mightymouse.brynmawr.edu>

On Tue, 2006-06-27 at 16:09 -0400, Grant Braught wrote:
> I'll be using Pyro in my AI course in the Fall and I've just begun work 
> on some homework exercises for the Vision unit.  What I want is for the 
> students to be able to write a brain that creates a FakeCamera and 
> points it to a directory containing a bunch of ppm image files that 
> would then be displayed and processed in much the same way the 
> FakeCamrea and the TutorialCamera work.  I saw the page on "Simulated 
> Vision Using FakeCamera" but it seems to be missing some information.  
> The second paragraph refers to two examples that would seem to be 
> exactly what I want but the examples don't seem to be there.
> 
> Any suggestions for or examples of how to create a FakeCamera using a 
> custom set of images within a brain would be greatly appreciated.

Grant, 

You should be able to make a fake camera device by putting something
like the following in a file:

# MyCamera.py ###################################################
from pyrobot.camera.fake import FakeCamera
from pyrobot.vision.cvision import VisionSystem
def INIT(robot):
    return {"camera": FakeCamera(pattern = "vision/blimp/hall-?.pbm",
                                 start = 1,
                                 stop = 9,
                                 interval = 1,
                                 visionSystem = VisionSystem())}
##################################################################

and loading this as a device (ie, by clicking on the "Device:" button,
or by selecting Load->Device from the menu).

The "?" marks the pattern where the number goes; use multiple question
marks to add zero padded numbers. For example, if you have:

/home/myid/mydir/image01.ppm
/home/myid/mydir/image02.ppm
/home/myid/mydir/image03.ppm
/home/myid/mydir/image04.ppm
/home/myid/mydir/image05.ppm

then pattern should be "/home/myid/mydir/image??.ppm", start = 1, stop =
5, and interval = 1 (more to skip some).

Currently, the ppm parser is very picky---it doesn't even like comments
at the top of the image file. That should be replaced with something
more robust, so your students don't have to edit the PPMs. We could even
allow gif, jpeg, etc. there. Does that sound useful?

Let us know if that doesn't work for you.

-Doug

> Thanks,
> Grant
> 
> 
> 
> _______________________________________________
> Pyro-users mailing list
> Pyro-users@emergent.brynmawr.edu
> http://emergent.brynmawr.edu/mailman/listinfo/pyro-users
> 
-- 
Douglas S. Blank       Computer Science
Assistant Professor    Bryn Mawr College
(610)526-6501          http://cs.brynmawr.edu/~dblank


From fox at macalester.edu  Wed Jun 28 12:23:11 2006
From: fox at macalester.edu (Susan Fox)
Date: Wed Jun 28 12:21:21 2006
Subject: [Pyro-users] Real pioneer robot, gripper, and camera devices..
Message-ID: <44A2ACEF.9050900@macalester.edu>

An HTML attachment was scrubbed...
URL: http://emergent.brynmawr.edu/pipermail/pyro-users/attachments/20060628/ae08ea73/attachment.htm
From dblank at brynmawr.edu  Wed Jun 28 17:50:19 2006
From: dblank at brynmawr.edu (Douglas S. Blank)
Date: Wed Jun 28 17:50:36 2006
Subject: [Pyro-users] Real pioneer robot, gripper, and camera devices..
In-Reply-To: <44A2ACEF.9050900@macalester.edu>
References: <44A2ACEF.9050900@macalester.edu>
Message-ID: <1151531419.2777.113.camel@mightymouse.brynmawr.edu>

On Wed, 2006-06-28 at 11:23 -0500, Susan Fox wrote:
> Hi,
> 
> I am running Pyro on a Pioneer 2 dx robot that has a gripper and a
> Sony PTZ camera on top.  I am hoping to be able to integrate gripper
> and  camera functions into an existing program I've written.  I know
> that I need a cable in order to control the PTZ functions of my
> camera, so I've been playing with the gripper commands as they are
> described in the Pyro Devices section, and also playing with the
> camera and vision processing stuff from the Computer Vision sections
> online.
> 
> First, the gripper:  I've found some confusions in working through the
> gripper stuff in the Pyro Devices section.  In places the text says to
> do things like "self.startDevice('gripper')", and I found that I
> needed to type "self.robot.startDevice('gripper')".  I assumed that
> those might be typos, or might depend on how I was connecting to a
> simulator/server and robot.  

Susan,

There are some methods and properties that obviously belong to a robot
(x and y for example). There are others that obviously belong to the
brain (step(), for example). Others, we have found, aren't clear to the
student what object they belong to. 

As an implementation detail, we have attached all devices to the robot.
We originally designed the method startDevice() to therefore be a method
of the robot class. But, we also made wrapper methods in the brain class
to pass on some of these requests the the robot. So, for example, in
pyrobot/brain/__init__.py you'll find something like:

class Brain:
   ...
   def startDevice(self, *args, **keywords):
      return self.robot.startDevice(*args, **keywords)
   ...

So, it shouldn't have mattered which one you called. However, I suspect
that there was a little problem: you hadn't loaded a brain yet. Planning
for that little oversight, we have a simple little object pre-defined as
"brain" and "self" so that you can type at the Pyro command line
'self.robot' or 'brain.robot' and it will refer to the robot even though
there is no brain loaded yet. Nice!

Except, that we forgot to put those wrappers in the pre-defined brain
object. I've updated the latest Pyro so either version will work, with
or without a loaded brain.

> I was running the Player Server, and using the Player.py robot,
> because I wanted to control the actual robot's devices.  I was able to
> control the gripper, except that the up and down movement commands
> didn't seem to work at all:  open() and close() worked perfectly, and
> deploy() and store() caused the gripper to open or shut as they
> should, but no vertical movement took place, not with deploy() and
> store(), and not with up() and down().  I wondered if there might be a
> bug in there somewhere.

Hmmm. deploy() and store() should also have vertical movements. It
sounds like a problem with your gripper to me, but I'll check it out on
our end too.

> Secondly, the camera:  Following the model in the Pyro Module Vision
> System section, I tried "pyrobot -r Test -d V4LCamera" and was able to
> see the live camera image from my robot's camera.  I had great fun
> figuring out the filtering commands. Using the Player Server and the
> Player.py robot client, I was able to add the camera by clicking on
> the "Devices"  button.  Is there a way to automate the loading of the
> camera somehow?

Sure, exactly the way that you started the gripper above:

self.robot.startDevice("V4LCamera")

BTW, capital letters in the device name mean that the device is coming
from the pyrobot/plugins/devices directory as external devices. Lower
case letters mean that the device is a "built-in" and the robot class
will take care of loading it. The V4LCamera (Video for Linux) can be
used with any robot on that computer.

Hope that helps,

-Doug

> Thanks,
> 
> Susan Fox
> -- 
> Susan Fox                                 Associate Professor
> Mathematics/Computer Science              Macalester College
> Office: (651) 696-6553                    Fax: (651) 696-6518
> fox@macalester.edu                        http://www.macalester.edu/~fox
> 
> _______________________________________________
> Pyro-users mailing list
> Pyro-users@emergent.brynmawr.edu
> http://emergent.brynmawr.edu/mailman/listinfo/pyro-users
-- 
Douglas S. Blank       Computer Science
Assistant Professor    Bryn Mawr College
(610)526-6501          http://cs.brynmawr.edu/~dblank


From braught at dickinson.edu  Thu Jun 29 08:40:41 2006
From: braught at dickinson.edu (Grant Braught)
Date: Thu Jun 29 08:41:06 2006
Subject: [Pyro-users] Custom FakeCamera Images
In-Reply-To: <1151443406.2777.34.camel@mightymouse.brynmawr.edu>
References: <c1a311c4f0115819198dab334dedaa0d@dickinson.edu>
	<1151443406.2777.34.camel@mightymouse.brynmawr.edu>
Message-ID: <acd0e6e6c64a0218593b6cb31ed012cc@dickinson.edu>

Doug,

Thanks, your suggestion worked like a charm!  I just wanted to post 
back to mention the great resource that I've come across for doing some 
vision activities that led to my original question.  The SIVAL Image 
Repository (http://www.cs.wustl.edu/~sg/accio/SIVAL.html) is a great 
collection of photos of a variety of objects in a wide range of 
settings.  Every object is photographed in 6 different locations and 
zoom (wide-angle or close-up) within each of 10 different settings.  
Perfect for object discrimination and object recognition tasks.

I've converted all of the images from jpg to ppm format and also 
written a script that produces custom device files for each image set 
so that they can be used as FakeCameras.  If anyone is interested in 
these files or the types of activities I'm planning with them, they can 
feel free to contact me.

Thanks again for you help!
Grant
_____________________________________________________________
Grant Braught					|	Phone: (717) 245-1401
Assoc. Prof. of Computer Science	|        Fax: (717) 245-1642
Dickinson College				|	EMail: braught@dickinson.edu
P.O. Box 1773					|	Web: www.dickinson.edu/~braught
Carlisle, PA 17013				|


On Jun 27, 2006, at 5:23 PM, Douglas S. Blank wrote:

> On Tue, 2006-06-27 at 16:09 -0400, Grant Braught wrote:
>> I'll be using Pyro in my AI course in the Fall and I've just begun 
>> work
>> on some homework exercises for the Vision unit.  What I want is for 
>> the
>> students to be able to write a brain that creates a FakeCamera and
>> points it to a directory containing a bunch of ppm image files that
>> would then be displayed and processed in much the same way the
>> FakeCamrea and the TutorialCamera work.  I saw the page on "Simulated
>> Vision Using FakeCamera" but it seems to be missing some information.
>> The second paragraph refers to two examples that would seem to be
>> exactly what I want but the examples don't seem to be there.
>>
>> Any suggestions for or examples of how to create a FakeCamera using a
>> custom set of images within a brain would be greatly appreciated.
>
> Grant,
>
> You should be able to make a fake camera device by putting something
> like the following in a file:
>
> # MyCamera.py ###################################################
> from pyrobot.camera.fake import FakeCamera
> from pyrobot.vision.cvision import VisionSystem
> def INIT(robot):
>     return {"camera": FakeCamera(pattern = "vision/blimp/hall-?.pbm",
>                                  start = 1,
>                                  stop = 9,
>                                  interval = 1,
>                                  visionSystem = VisionSystem())}
> ##################################################################
>
> and loading this as a device (ie, by clicking on the "Device:" button,
> or by selecting Load->Device from the menu).
>
> The "?" marks the pattern where the number goes; use multiple question
> marks to add zero padded numbers. For example, if you have:
>
> /home/myid/mydir/image01.ppm
> /home/myid/mydir/image02.ppm
> /home/myid/mydir/image03.ppm
> /home/myid/mydir/image04.ppm
> /home/myid/mydir/image05.ppm
>
> then pattern should be "/home/myid/mydir/image??.ppm", start = 1, stop 
> =
> 5, and interval = 1 (more to skip some).
>
> Currently, the ppm parser is very picky---it doesn't even like comments
> at the top of the image file. That should be replaced with something
> more robust, so your students don't have to edit the PPMs. We could 
> even
> allow gif, jpeg, etc. there. Does that sound useful?
>
> Let us know if that doesn't work for you.
>
> -Doug
>
>> Thanks,
>> Grant
>>
>>
>>
>> _______________________________________________
>> Pyro-users mailing list
>> Pyro-users@emergent.brynmawr.edu
>> http://emergent.brynmawr.edu/mailman/listinfo/pyro-users
>>
> -- 
> Douglas S. Blank       Computer Science
> Assistant Professor    Bryn Mawr College
> (610)526-6501          http://cs.brynmawr.edu/~dblank
>

From dblank at brynmawr.edu  Thu Jun 29 11:43:38 2006
From: dblank at brynmawr.edu (Douglas S. Blank)
Date: Thu Jun 29 11:43:32 2006
Subject: [Pyro-users] Pyro and Vision
In-Reply-To: <acd0e6e6c64a0218593b6cb31ed012cc@dickinson.edu>
References: <c1a311c4f0115819198dab334dedaa0d@dickinson.edu>
	<1151443406.2777.34.camel@mightymouse.brynmawr.edu>
	<acd0e6e6c64a0218593b6cb31ed012cc@dickinson.edu>
Message-ID: <1151595818.3678.23.camel@mightymouse.brynmawr.edu>

There has been a lot of activity from colleagues interested in vision
and Pyro lately. Grant has kindly added to these efforts by sending his
instructions on converting the SIVAL collection into a format that can
easily be used by Pyro, even the Pyro Live CD. Very handy!

I've added this information on the wiki page:

http://pyrorobotics.org/?page=Image_20Processing_20Collections

along with some initial information on exploring stereo depth perception
with Pyro. This now appears off of the Curriculum -> Pyro Module
Computer Vision page.

If anyone would like to add/correct information on this page (or any
PyroRobotics.org page) please feel free to do so.

Thanks again, Grant!

-Doug

On Thu, 2006-06-29 at 08:40 -0400, Grant Braught wrote:
> Doug,
> 
> Thanks, your suggestion worked like a charm!  I just wanted to post 
> back to mention the great resource that I've come across for doing some 
> vision activities that led to my original question.  The SIVAL Image 
> Repository (http://www.cs.wustl.edu/~sg/accio/SIVAL.html) is a great 
> collection of photos of a variety of objects in a wide range of 
> settings.  Every object is photographed in 6 different locations and 
> zoom (wide-angle or close-up) within each of 10 different settings.  
> Perfect for object discrimination and object recognition tasks.
> 
> I've converted all of the images from jpg to ppm format and also 
> written a script that produces custom device files for each image set 
> so that they can be used as FakeCameras.  If anyone is interested in 
> these files or the types of activities I'm planning with them, they can 
> feel free to contact me.
> 
> Thanks again for you help!
> Grant
> _____________________________________________________________
> Grant Braught					|	Phone: (717) 245-1401
> Assoc. Prof. of Computer Science	|        Fax: (717) 245-1642
> Dickinson College				|	EMail: braught@dickinson.edu
> P.O. Box 1773					|	Web: www.dickinson.edu/~braught
> Carlisle, PA 17013				|
> 
> 
> On Jun 27, 2006, at 5:23 PM, Douglas S. Blank wrote:
> 
> > On Tue, 2006-06-27 at 16:09 -0400, Grant Braught wrote:
> >> I'll be using Pyro in my AI course in the Fall and I've just begun 
> >> work
> >> on some homework exercises for the Vision unit.  What I want is for 
> >> the
> >> students to be able to write a brain that creates a FakeCamera and
> >> points it to a directory containing a bunch of ppm image files that
> >> would then be displayed and processed in much the same way the
> >> FakeCamrea and the TutorialCamera work.  I saw the page on "Simulated
> >> Vision Using FakeCamera" but it seems to be missing some information.
> >> The second paragraph refers to two examples that would seem to be
> >> exactly what I want but the examples don't seem to be there.
> >>
> >> Any suggestions for or examples of how to create a FakeCamera using a
> >> custom set of images within a brain would be greatly appreciated.
> >
> > Grant,
> >
> > You should be able to make a fake camera device by putting something
> > like the following in a file:
> >
> > # MyCamera.py ###################################################
> > from pyrobot.camera.fake import FakeCamera
> > from pyrobot.vision.cvision import VisionSystem
> > def INIT(robot):
> >     return {"camera": FakeCamera(pattern = "vision/blimp/hall-?.pbm",
> >                                  start = 1,
> >                                  stop = 9,
> >                                  interval = 1,
> >                                  visionSystem = VisionSystem())}
> > ##################################################################
> >
> > and loading this as a device (ie, by clicking on the "Device:" button,
> > or by selecting Load->Device from the menu).
> >
> > The "?" marks the pattern where the number goes; use multiple question
> > marks to add zero padded numbers. For example, if you have:
> >
> > /home/myid/mydir/image01.ppm
> > /home/myid/mydir/image02.ppm
> > /home/myid/mydir/image03.ppm
> > /home/myid/mydir/image04.ppm
> > /home/myid/mydir/image05.ppm
> >
> > then pattern should be "/home/myid/mydir/image??.ppm", start = 1, stop 
> > =
> > 5, and interval = 1 (more to skip some).
> >
> > Currently, the ppm parser is very picky---it doesn't even like comments
> > at the top of the image file. That should be replaced with something
> > more robust, so your students don't have to edit the PPMs. We could 
> > even
> > allow gif, jpeg, etc. there. Does that sound useful?
> >
> > Let us know if that doesn't work for you.
> >
> > -Doug
> >
> >> Thanks,
> >> Grant
> >>
> >>
> >>
> >> _______________________________________________
> >> Pyro-users mailing list
> >> Pyro-users@emergent.brynmawr.edu
> >> http://emergent.brynmawr.edu/mailman/listinfo/pyro-users
> >>
> > -- 
> > Douglas S. Blank       Computer Science
> > Assistant Professor    Bryn Mawr College
> > (610)526-6501          http://cs.brynmawr.edu/~dblank
> >
> 
-- 
Douglas S. Blank       Computer Science
Assistant Professor    Bryn Mawr College
(610)526-6501          http://cs.brynmawr.edu/~dblank