Re: [Playerstage-users] Points not set properly in ranger when laser-like devices are used

[Fred L. <ff...@ab...>]

On Sunday 11 March 2012 18:57:24 Rich Mattes wrote:
> On 03/10/2012 01:47 PM, Fred Labrosse wrote:
> > On Friday 09 March 2012 15:17:11 Fred Labrosse wrote:
> >> All,
> >> 
> >> I just noticed that laser-like rangers don't have their points set
> >> properly. The reason is that in playerc_ranger_calculate_points(),
> >> the given device has 0 for min_angle and angular_res.  the outcome is
> >> that the value for px is the range and the value for py is always 0.
> >> 
> >> Any idea where these are set?  Have not had time to delve much yet.  I
> >> have requested the geometry, following me discovering something
> >> similar with sonars.
> >> 
> >> Any hint would be much appreciated.
> > 
> > Same problem as one I had before, about geometry of the ranger.  An
> > explicit call to rangerProxy::RequestConfigure() is needed!  As for the
> > "missing" RequestGeom(), should that not be automatically called?
> > 
> > Cheers,
> > 
> > Fred
> 
> RequestConfigure() and RequestGeom() are now called within the
> RangerProxy constructor (committed to trunk this afternoon).  It seems
> to be working just fine in my testing with Stage and playerprint, let me
> know if you have any issues.

I'll try that.  Thanks for doing that.

I noticed that the coordinate system for 1-element and multiple-element 
rangers are different.

I have a pioneer (simulated) with 1 front and 1 back ranger (sonar) and 
another robot (simulated and real) with front and back rangers (laser).  In 
playerv (and stage when simulated) the rangers point the correct way (front 
and back) but I get points that are at the back for the sonar but at the front 
for the laser.  Both playerv and stage do not use points but compute the 
points themselves, while I use points.

My understanding is that points are in the robot coordinate system, not the 
device's.  Am I right?  It is clear that the ranger function 
playerc_ranger_calculate_points() does take into account the pose of 
individual elements, but not that of the device (or the element) if there is a 
single element.

I'm happy to work on that, if someone can confirm which coordinate system the 
points are supposed to be in.

Cheers,

Fred


> 
> Rich
> 
> 
> ----------------------------------------------------------------------------
> -- Virtualization & Cloud Management Using Capacity Planning
> Cloud computing makes use of virtualization - but cloud computing
> also focuses on allowing computing to be delivered as a service.
> http://www.accelacomm.com/jaw/sfnl/114/51521223/
> _______________________________________________
> Playerstage-users mailing list
> Pla...@li...
> https://lists.sourceforge.net/lists/listinfo/playerstage-users