[robocup-small] robocup-small Digest, Vol 55, Issue 4

Stefan Zickler zickler at gmail.com
Thu Aug 13 09:55:02 EDT 2009


Hi All,

thanks for the discussion on this. There is a point I would like to
add besides the ones mentioned by RoboFEI:

One of the main goals and benefits of SSL-Vision is to finally ease
the ridiculous setup and organizational struggles of the SSL. By
making SSL-Vision mandatory (and providing a single set of cameras + a
vision server per field), teams will be able to switch fields easily
(like in any other league), thus allowing a more flexible game
schedule. Furthermore, time between games can be used for actual work
on A.I. and robot testing, instead of calibrating cameras and chasing
ladders around. Scalability to more cameras is another advantage, as
RoboFEI mentioned.
If we were to make the use of SSL-Vision optional then we would
effectively keep the status quo, having non-participating teams still
be bound to their fields and blocking the field by putting up their
own cameras/equipment.

Of course nobody wants to block innovation. This question has already
been carefully addressed during the previous road-map discussions and
during development of SSL-Vision (the SSL-Vision paper of the 2009
proceedings also touched upon this topic). That is why, as RoboFEI
pointed out, SSL-Vision is an open-source project, and is very modular
in its design. If you or your team think that you have an improved
vision algorithm, or an additional feature that SSL-Vision is missing
then this can be easily evaluated and added to the codebase. In fact,
we are thankful for any additional developers. :-)

Best Regards,
Stefan

On Thu, Aug 13, 2009 at 8:24 AM, <jgurzoni at yahoo.com.br> wrote:
> Ravish, greetings.
> My first gut reaction when I heard of mandatory was also similar to yours, but if you read the SSL roadmap document you may change your mind. With a standard vision, leaving aside the argument that new teams would benefit from the availability of such system (as even as optional this benefit holds), you still have other good reasons:
> - it is easier to increase the no. of cameras and then be able to increase the field size. Right now its already a mess to have all the cameras of all teams.
> - little advance was made scientifically with vision systems. proof of that is that most teams have similar solutions today. Also, the system doesn't give you just everything but rather the basics. You still have to work on tracking and etc.
> - you can still contribute designing plug-ins to the system, as we are doing in RoboFEI team. We're developing a neural network based color calibration system. The ssl-vision developers really made it modular and easy to add plug-ins.
> - you can contribute to the ssl-vision project as well. Its an open source.
>
>
>
>
> Cheers
> Junior
> RoboFEI Team
>
> Sent via BlackBerry from T-Mobile
>
> -----Original Message-----
> From: Ravish Sharma <ravish.hacker at gmail.com>
>
> Date: Thu, 13 Aug 2009 13:55:07
> To: <robocup-small at cc.gatech.edu>
> Subject: Re: [robocup-small] robocup-small Digest, Vol 55, Issue 4
>
>
> _______________________________________________
> robocup-small mailing list
> robocup-small at cc.gatech.edu
> https://lists.cc.gatech.edu/mailman/listinfo/robocup-small
>
> _______________________________________________
> robocup-small mailing list
> robocup-small at cc.gatech.edu
> https://lists.cc.gatech.edu/mailman/listinfo/robocup-small
>


More information about the robocup-small mailing list