[Ros-kinect] Timing Calibration

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

[Ros-kinect] Timing Calibration

Tully Foote
Hi,

Has anyone been looking into characterizing the timing delays for the Kinect data?  I've noticed that there are some recofigure options for changing the timestamps.  Has anyone been using these sliders and if so have they developed ways to characterize the delay?  I've played with the sliders and observed close to -0.12 as the optimal offset for the depth data. 

Tully

--
Tully Foote
Systems Engineer
Willow Garage, Inc.
[hidden email]
(650) 475-2827

_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect
Reply | Threaded
Open this post in threaded view
|

Re: [Ros-kinect] Timing Calibration

Suat Gedikli
Hi Tully,

ask Eitan, Wim and Stu. I added the time-offset feature for them.
-Suat

On Tue, May 3, 2011 at 12:19 AM, Tully Foote <[hidden email]> wrote:
Hi,

Has anyone been looking into characterizing the timing delays for the Kinect data?  I've noticed that there are some recofigure options for changing the timestamps.  Has anyone been using these sliders and if so have they developed ways to characterize the delay?  I've played with the sliders and observed close to -0.12 as the optimal offset for the depth data. 

Tully

--
Tully Foote
Systems Engineer
Willow Garage, Inc.
[hidden email]
<a href="tel:%28650%29%20475-2827" value="+16504752827" target="_blank">(650) 475-2827

_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect



_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect
Suat
Reply | Threaded
Open this post in threaded view
|

Re: [Ros-kinect] Timing Calibration

Eitan Marder-Eppstein
Hey all,

Sorry its taken me so long to chime in on this. I believe that I needed the time-offset feature for running navigation on a PR2 with the Kinect. The problem was that the driver was stamping clouds at the time that they were received from the Kinect over usb, but I needed them stamped at the time that they were actually taken by the Kinect because I was tilting the head of the PR2. The time that the data was taking to go over USB was actually screwing up localization because point clouds were not being transformed correctly.

An offset of around -0.12 sounds about right, but its been awhile since I played with it. I do remember that the offset was more-or-less fixed.

Hope this helps,

Eitan

On Tue, May 3, 2011 at 1:04 AM, Suat Gedikli <[hidden email]> wrote:
Hi Tully,

ask Eitan, Wim and Stu. I added the time-offset feature for them.
-Suat

On Tue, May 3, 2011 at 12:19 AM, Tully Foote <[hidden email]> wrote:
Hi,

Has anyone been looking into characterizing the timing delays for the Kinect data?  I've noticed that there are some recofigure options for changing the timestamps.  Has anyone been using these sliders and if so have they developed ways to characterize the delay?  I've played with the sliders and observed close to -0.12 as the optimal offset for the depth data. 

Tully

--
Tully Foote
Systems Engineer
Willow Garage, Inc.
[hidden email]
<a href="tel:%28650%29%20475-2827" value="+16504752827" target="_blank">(650) 475-2827

_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect



_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect



_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect
Reply | Threaded
Open this post in threaded view
|

Re: [Ros-kinect] Timing Calibration

Tully Foote
Eitan,

Navigation is exactly my application.  Did you do anything automated for finding this offset or was it more trial and error with the slider?

Tully

On Tue, May 3, 2011 at 10:15 AM, Eitan Marder-Eppstein <[hidden email]> wrote:
Hey all,

Sorry its taken me so long to chime in on this. I believe that I needed the time-offset feature for running navigation on a PR2 with the Kinect. The problem was that the driver was stamping clouds at the time that they were received from the Kinect over usb, but I needed them stamped at the time that they were actually taken by the Kinect because I was tilting the head of the PR2. The time that the data was taking to go over USB was actually screwing up localization because point clouds were not being transformed correctly.

An offset of around -0.12 sounds about right, but its been awhile since I played with it. I do remember that the offset was more-or-less fixed.

Hope this helps,

Eitan


On Tue, May 3, 2011 at 1:04 AM, Suat Gedikli <[hidden email]> wrote:
Hi Tully,

ask Eitan, Wim and Stu. I added the time-offset feature for them.
-Suat

On Tue, May 3, 2011 at 12:19 AM, Tully Foote <[hidden email]> wrote:
Hi,

Has anyone been looking into characterizing the timing delays for the Kinect data?  I've noticed that there are some recofigure options for changing the timestamps.  Has anyone been using these sliders and if so have they developed ways to characterize the delay?  I've played with the sliders and observed close to -0.12 as the optimal offset for the depth data. 

Tully

--
Tully Foote
Systems Engineer
Willow Garage, Inc.
[hidden email]
<a href="tel:%28650%29%20475-2827" value="+16504752827" target="_blank">(650) 475-2827

_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect



_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect



_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect




--
Tully Foote
Systems Engineer
Willow Garage, Inc.
[hidden email]
(650) 475-2827

_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect
Reply | Threaded
Open this post in threaded view
|

Re: [Ros-kinect] Timing Calibration

Eitan Marder-Eppstein

Tully,

Trial and error looking at the ground plane with the kinect being tilted. When I got one more-or-less consistent plane, I called things good enough to go with.

Hope this helps,

Eitan

On May 3, 2011 3:49 PM, "Tully Foote" <[hidden email]> wrote:
> Eitan,
>
> Navigation is exactly my application. Did you do anything automated for
> finding this offset or was it more trial and error with the slider?
>
> Tully
>
> On Tue, May 3, 2011 at 10:15 AM, Eitan Marder-Eppstein <
> [hidden email]> wrote:
>
>> Hey all,
>>
>> Sorry its taken me so long to chime in on this. I believe that I needed the
>> time-offset feature for running navigation on a PR2 with the Kinect. The
>> problem was that the driver was stamping clouds at the time that they were
>> received from the Kinect over usb, but I needed them stamped at the time
>> that they were actually taken by the Kinect because I was tilting the head
>> of the PR2. The time that the data was taking to go over USB was actually
>> screwing up localization because point clouds were not being transformed
>> correctly.
>>
>> An offset of around -0.12 sounds about right, but its been awhile since I
>> played with it. I do remember that the offset was more-or-less fixed.
>>
>> Hope this helps,
>>
>> Eitan
>>
>>
>> On Tue, May 3, 2011 at 1:04 AM, Suat Gedikli <[hidden email]>wrote:
>>
>>> Hi Tully,
>>>
>>> ask Eitan, Wim and Stu. I added the time-offset feature for them.
>>> -Suat
>>>
>>> On Tue, May 3, 2011 at 12:19 AM, Tully Foote <[hidden email]>wrote:
>>>
>>>> Hi,
>>>>
>>>> Has anyone been looking into characterizing the timing delays for the
>>>> Kinect data? I've noticed that there are some recofigure options for
>>>> changing the timestamps. Has anyone been using these sliders and if so have
>>>> they developed ways to characterize the delay? I've played with the sliders
>>>> and observed close to -0.12 as the optimal offset for the depth data.
>>>>
>>>> Tully
>>>>
>>>> --
>>>> Tully Foote
>>>> Systems Engineer
>>>> Willow Garage, Inc.
>>>> [hidden email]
>>>> (650) 475-2827
>>>>
>>>> _______________________________________________
>>>> Ros-kinect mailing list
>>>> [hidden email]
>>>> https://code.ros.org/mailman/listinfo/ros-kinect
>>>>
>>>>
>>>
>>> _______________________________________________
>>> Ros-kinect mailing list
>>> [hidden email]
>>> https://code.ros.org/mailman/listinfo/ros-kinect
>>>
>>>
>>
>> _______________________________________________
>> Ros-kinect mailing list
>> [hidden email]
>> https://code.ros.org/mailman/listinfo/ros-kinect
>>
>>
>
>
> --
> Tully Foote
> Systems Engineer
> Willow Garage, Inc.
> [hidden email]
> (650) 475-2827

_______________________________________________
Ros-kinect mailing list
[hidden email]
https://code.ros.org/mailman/listinfo/ros-kinect