[ros-users] Detecting Glass Walls

[Dejan Pangercic]

Hi there,
I've done some literature research but mostly on detection of
transparent objects. Here are some pointers:
http://www.goneneren.com/publications
http://www.robotic.de/fileadmin/robotic/hillenbd/Hillenbrand_08.pdf
http://www.willowgarage.com/additive-latent-feature-model-transparent-object-recognition

I plan to attack this problem in the next half a year and I am thus
very willing to hear about from the rest of community.

cheers, D.
On Fri, Apr 2, 2010 at 7:03 PM, Caroline Pantofaru
<pantofaru at willowgarage.com> wrote:
> I haven't done a literature search, but off the top of my head I would be
> tempted to look at the ground (if you can see it) to provide an indication
> that there might be a wall. Sometimes there's a bit of wall beneath the
> glass, or at least the flooring type/color/pattern will change. It's a hack,
> but it might help.
>
> On Fri, Apr 2, 2010 at 9:59 AM, Radu Bogdan Rusu <rusu at willowgarage.com>
> wrote:
>>
>> Yeah, plus they heavily depend on how the light reflects, angles, etc.
>> Reflections are evil ;) I haven't seen too much
>> work on actually getting stereo from that. Is there any work on
>> artificially generating reflections by manipulating
>> light (a PR2 with a flashlight moving around would be cool :) )?
>>
>> Cheers,
>> Radu.
>>
>> Kurt Konolige wrote:
>> > Don't forget that reflections will be seen by stereo as behind the
>> > glass...kk
>> >
>> > On Fri, Apr 2, 2010 at 9:44 AM, Radu Bogdan Rusu <rusu at willowgarage.com>
>> > wrote:
>> >> Hi Eric,
>> >>
>> >> Ouch, glass walls. No, unfortunately there isn't a lot that can be done
>> >> with LIDAR sensors there. You can try to use
>> >> reflections to generate stereo data and hope for the best, but that's
>> >> not going to work everytime. You are right,
>> >> employing sonars might help. Did you check the literature to see what
>> >> other people are using to deal with glass?
>> >>
>> >> Cheers,
>> >> Radu.
>> >>
>> >> Eric Perko wrote:
>> >>> Hey all,
>> >>>
>> >>> After looking at the wiki for ROS sonar packages, I started wondering.
>> >>> Currently, the robots I work on have issues detecting glass walls, as
>> >>> their primary range sensor is a SICK LMS291, which pings right through
>> >>> the glass (obviously). To solve this for a robotic tour guide
>> >>> application, we are looking into sonars.
>> >>>
>> >>> So my question is, what are some approaches that you use to detect
>> >>> things such as glass walls that both LIDARs and cameras might have
>> >>> trouble with? Can the PR2 see glass walls and, if so, how?
>> >>>
>> >>> Thanks
>> >>> Eric Perko
>> >>> wisesage5001 at gmail.com
>> >>> eric.perko at case.edu
>> >>> mobilerobots.case.edu
>> >>> _______________________________________________
>> >>> ros-users mailing list
>> >>> ros-users at code.ros.org
>> >>> https://code.ros.org/mailman/listinfo/ros-users
>> >> --
>> >> | Radu Bogdan Rusu | http://rbrusu.com/
>> >> _______________________________________________
>> >> ros-users mailing list
>> >> ros-users at code.ros.org
>> >> https://code.ros.org/mailman/listinfo/ros-users
>> >>
>>
>> --
>> | Radu Bogdan Rusu | http://rbrusu.com/
>> _______________________________________________
>> ros-users mailing list
>> ros-users at code.ros.org
>> https://code.ros.org/mailman/listinfo/ros-users
>
>
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>
>



-- 
MSc. Dejan Pangercic
PhD Student/Researcher
Computer Science IX
Technische Universität München
Telephone: +49 (89) 289-17780
E-Mail: dejan.pangercic at in.tum.de
WWW: http://ias.cs.tum.edu/people/pangercic