[Rtk-users] Geometric calibration

[Simon Rit]

Hi,
Enclosed. Column 1 is platform angle, 2 is bb height, 3 is the u coordinate
(in pixel) and 4 is the v coordinate (in pixel too).
I had approximately measured the source-to-detector (270) and
source-to-isocenter (210) distances but they were still optimized to have a
better assessment.
Simon

On Thu, Nov 24, 2016 at 7:34 AM, Sepp de Raedt <sdr at nrtxray.com> wrote:

> Hi Simon,
>
> Thanks for the script. I was wondering if you could also share the csv or
> a dummy file? I'm unsure of what the second column contains. An offset
> distance to the rotation axis? Or the height?
>
> Which distance was known in your example? From the code it looks like the
> distance to the rotation axis?
>
> Sepp
>
> On 22 Nov 2016, at 08.01, Simon Rit <simon.rit at creatis.insa-lyon.fr>
> wrote:
>
> Hi Sepp,
> I did do a calibration script on a micro CT platform. The idea was that I
> had a bb taken from 4 different view points at 4 different heights with a
> known distance. Their projected position was segmented beforehand on the
> projections and I tried to match the projections by optimizing the
> calibration (which we see in before.pdf and after.pdf). I don't think it's
> a good script, e.g., because I tried to find 8 parameters when 7 are
> sufficient for my system (see, e.g., this paper
> <http://iopscience.iop.org/article/10.1088/0031-9155/45/11/327/meta;jsessionid=8DD0F1160098FBC65BB161A0257FBA75.c3.iopscience.cld.iop.org#>).
> But this gives you a starting point to develop your own script which will
> hopefully be better. Please share if you do something from it!
> Thanks,
> Simon
>
> On Sat, Nov 19, 2016 at 8:51 PM, Sepp de Raedt <sdr at nrtxray.com> wrote:
>
>> Hi RTK users,
>>
>> I'm interested in reconstructing images acquired by a fluoroscopy system
>> rotating around a knee phantom. Unfortunately, I've been unsuccessful so
>> far. I recognize some of the phantom in the reconstructed image, but it
>> contains double contours and the shape is deformed.
>>
>> I think the issue might be incorrect geometry specification. I have
>> images also including a phantom containing markers for which we know the
>> coordinates, which should allow us to calibrate the system. On the wiki on
>> image quality, I saw that some users had developed scripts to do the
>> calibration? Would someone be willing to share it?
>>
>> Kind regards,
>> Sepp
>>
>> ______________________________________________________________________
>> This email has been scanned by the Symantec Email Security.cloud service.
>> For more information please visit http://www.symanteccloud.com
>> ______________________________________________________________________
>> _______________________________________________
>> Rtk-users mailing list
>> Rtk-users at public.kitware.com
>> http://public.kitware.com/mailman/listinfo/rtk-users
>>
>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please visit http://www.symanteccloud.com
> ______________________________________________________________________
>
> <after.pdf>
>
> <before.pdf>
>
> <calib.py>
>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please visit http://www.symanteccloud.com
> ______________________________________________________________________
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/rtk-users/attachments/20161124/6b6cfa85/attachment-0010.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: CentroidPositions_H5_to_H25_90degstep.csv
Type: text/csv
Size: 321 bytes
Desc: not available
URL: <http://public.kitware.com/pipermail/rtk-users/attachments/20161124/6b6cfa85/attachment-0010.csv>