[ros-users] pcl inconsistencies

Ken Conley kwc at willowgarage.com
Wed Aug 11 18:41:36 UTC 2010


Hi Jared,

You should keep using cturtle. It will continue to receive updates to pcl.

'latest' is no longer being developed -- our new bleeding edge tree
will be called 'unstable', but we are not actively developing on it
yet.

 - Ken

On Wed, Aug 11, 2010 at 11:36 AM, Jared Marshall Glover <jglov at mit.edu> wrote:
> Hi Radu,
>
> I'm still using cturtle, not latest.  Once we finish our book grabbing
> demo this
> week, I'll update to latest and give it a shot =)  Thanks for fixing it so
> quickly.  By the way, opencv seems to have a similar problem of using degrees
> instead of radians (e.g. in cv::getRotationMatrix2D).  Do you know if
> there are
> plans to adopt the same SI units convention in opencv anytime soon?
>
> -Jared
>
>
>
> Quoting Radu Bogdan Rusu <rusu at willowgarage.com>:
>
>> Hi Jared,
>>
>> I pushed 0.2.4 out. Can you please let me know if that fixed the problem?
>>
>> Thanks,
>> Radu.
>>
>> On 08/10/2010 08:01 PM, Jared Marshall Glover wrote:
>>> Hi Radu,
>>>
>>> Yes, after a quick grep for "180" in pcl/src, it looks like the problem
>>> is just
>>> in sac_segmentation.hpp, lines 153, 170, 294, and 318. But you'll also
>>> need to
>>> change any code (including on the wiki tutorials) that uses those models
>>> for
>>> segmentation.
>>>
>>> -Jared
>>>
>>>
>>>
>>> Quoting Radu Bogdan Rusu <rusu at willowgarage.com>:
>>>
>>>> Hi Jared,
>>>>
>>>> Is this just in sac_segmentation.hpp (153, 170), or did you find it
>>>> anywhere else?
>>>>
>>>> Thanks,
>>>> Radu.
>>>>
>>>> On 08/10/2010 02:17 PM, Jared Marshall Glover wrote:
>>>>> Hi Radu,
>>>>>
>>>>> Ok I created a ticket. But it may be a larger issue than just
>>>>> consistency in
>>>>> SACSegmentation::setEpsAngle(). Also, you don't necessarily want to
>>>>> break
>>>>> existing code without warning. Maybe you should decide on a unit
>>>>> system and
>>>>> put a clear indication in the documentation that as of version XXX all
>>>>> units
>>>>> will be standardized?
>>>>>
>>>>> -Jared
>>>>>
>>>>>
>>>>> Quoting Radu Bogdan Rusu <rusu at willowgarage.com>:
>>>>>
>>>>>> Hi Jared,
>>>>>>
>>>>>> oops. Can you please send me a patch (ticket works too)?
>>>>>>
>>>>>> Thanks,
>>>>>> Radu.
>>>>>>
>>>>>> On 08/10/2010 02:05 PM, Jared Marshall Glover wrote:
>>>>>>> There seems to be an inconsistency in PCL (at least in cturtle...I
>>>>>>> haven't
>>>>>>> checked latest), in that some methods take angles in radians and
>>>>>>> some in
>>>>>>> degrees. For example, SACSegmentation::setEpsAngle() accepts
>>>>>>> degrees for
>>>>>>> SACMODEL_ORIENTED_PLANE, but radians for SACMODEL_ORIENTED_LINE.
>>>>>>> Perhaps there
>>>>>>> should be a single convention throughout PCL to avoid confusion? It's
>>>>>>> already
>>>>>>> caused some headaches around here =)
>>>>>>>
>>>>>>> -Jared
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> ros-users mailing list
>>>>>>> ros-users at code.ros.org
>>>>>>> https://code.ros.org/mailman/listinfo/ros-users
>>>>>>
>>>>>> --
>>>>>> | Radu Bogdan Rusu | http://rbrusu.com/
>>>>>>
>>>>>
>>>>>
>>>>
>>>> --
>>>> | Radu Bogdan Rusu | http://rbrusu.com/
>>>>
>>>
>>>
>>
>> --
>> | Radu Bogdan Rusu | http://rbrusu.com/
>>
>
>
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>



More information about the ros-users mailing list