[robocup-humanoid] ?: One question about the position ofImagesensor.

[Rodrigo Guerra]

Just a note for the future: If you think carefully cameras equipped
with omni-mirrors are as much human-like as projective cameras are.
The limited "field-of-view" is a human constraint that, in my view,
should be introduced sooner or later. I believe posting in this way
makes things look much less "scary" to those whose systems are
currently based on omni-cameras -- e.g. placing a black tape reducing
the field of view of such cameras would make the robot compliant to
the rules right away.

On 2/4/07, Chew Chee Meng <mpeccm at nus.edu.sg> wrote:
> Dear Oskar
>
> Our team had refrained from putting the camera below shoulder and spend a lot of effort to comply to this rule. It is a bit of a challenge especially with omnivision at the same place.  We were not aware that your team used a chest camera in 2006 and the fact that your team was still allowed to compete should be considered to be a one-off waiver or tweaking of rule.
>
> Btw, as omnivision is still approved for 2007 event, pls do not create more confusion by saying that it is non human-like.  We have to first define what we mean by human-like.  In fact, our team had learned a lot of the omnivDear Chew Chee,
>
> On Sat, 3 Feb 2007, Chew Chee Meng wrote:
>
> > Dear Oskar
> >
> > Our main point is that if we want to put it down explicitly in the rule,
> > it should be enforced.  Otherwise, each team will start to breach the
> > rule and expect the organiser to waive it.  Also, other teams may use
> > the same case to justify their breach of rules, e.g. claiming that XX
> > team also did it last year.  And we think it is unfair to teams who
>
> I think if the rule in last year had allowed it (not matter what it is)
> and the corersponding rule had not changed it is obviously allowed this
> year also.
>
> Perhaps we are not talking exactly about the same.
> My concrete point is:
> We are using (almost) the same robot design in 2007 which we used in 2006
> and which was allowed in 2006 (with a camera in the chest and a camera
> in the head).
> The corresponding wording in the rule had also not changed.
>
> To me it seems that you want to ban the use of our chest camera now.
> In principle we can put our chest camera as well in the head
> but this needs some time for hardware modification.
> Such a request at this time would be higly unfair
> to any team which has already prepared their robots and software for
> the qualification under the above mentioned conditions.
>
> As I remember your robot had been using omnivision in 2006 which is much
> less human like than a directed camera in the chest as our robot never can
> see a ball in the back of it and must plan to search for the ball
> by coordinating vision and locomotion. This needs time
> and complex algorithms to do it efficiently. Therefore, more human
> like directed vision is a significant disadvantage to super human
> omnivision.
>
> Thank you very much for your understanding.
> With best regards,
> Oskar
>
> --
> Prof. Dr. Oskar von Stryk           E-Mail: stryk(at)sim.tu-darmstadt.de
> Simulation and Systems Optimization Phone:  ++49 (0) 6151-16-2513
> Technische Universitaet Darmstadt   Fax:    ++49 (0) 6151-16-6648
> Hochschulstr. 10                    http://www.sim.tu-darmstadt.de
> D-64289 Darmstadt, Germany
> _______________________________________________
> robocup-humanoid mailing list
> robocup-humanoid at mailman.cc.gatech.edu
> https://mailman.cc.gatech.edu/mailman/listinfo/robocup-humanoid
>


-- 
Rodrigo da Silva Guerra
PhD Student

Department of Adaptive Machine Systems
Graduate School of Engineering
Osaka University - Japan