[Qgis-user] Re: Unusual application of QGIS

[Alex Mandel]

You should also consider ImageJ which is somewhat standard in the
MicroBio world for this kind of stuff. It has all sorts of cool
auto-stiching and referencing plugins for your type of work. You also
wouldn't have to worry about projections then.

Of course as people have pointed out GIS can be used for your use case,
and I have indeed stiched paper scans of images (non-earth related) to
each other in a local coordinate system.

Enjoy,
Alex

On 05/08/2012 07:22 AM, Gerhardus Geldenhuis wrote:
> Thanks for the reply
> 
> On 8 May 2012 12:11, Rainer M Krug <r.m.krug at gmail.com> wrote:
> 
>>
>>
>>> I do EMPA ( Electron probe micro-analyzer ) of microscope slides. When
>> it is in the device I
>>> can note down x,y,z co-ordinates when taking a photo or doing a sample
>> analysis. The problem is
>>> when you remove the slide and put it back in again these previous x,y
>> and z components are now
>>> meaningless because the device that holds the slide is not accurate
>> enough to achieve the same
>>> positioning again. I was thinking that if I had known reference points
>> on the slide I could
>>> "geo-reference" all pictures and scans I did for a specific session and
>> then when I have follow
>>> up photos and point analysis data, taken in a separate session, I could
>> "geo-reference these
>>> and use QGIS to map out all the photos and pin point analysis as
>> different layers making it
>>> very easy to correlate data. As long as I have the co-ordinates I should
>> then be able to add
>>> any other microscope ( eg. petrographic images ) data as a separate
>> layer and match it up.
>>
>>
>> Sure - you can "geo-reference" whatever image you have against whatever
>> system. In your case, yo
>> should be choosing the right co-ordinate system and the right projection.
>> This is something I
>> can't help you with at the moment, but it should be easy enough to find
>> out. If in doubt, ask
>> specifically for this again, as it is of uttermost importance for
>> measuring distances, areas,
>> calculations and so on.
>>
> 
> All true, but at least it sounds possible. The images I am mapping is
> square so another thing to consider in my co-ordinate and projection
> system.
> 
> 
>>
>> One definitely has to be aware, that the space of the GIS, which is
>> usually the surface of the
>> earth (or a section of it) can be anything. It can simply be a piece of
>> paper or, in your case,
>> the area *flat* area which is seen in the EMPA. And you want to reference
>> your slide to the area,
>> so that a specific point of the slide is always at the same location -
>> look at it as satellite
>> pictures: the earth is your slide, and the satellite is your EMPA (OK -
>> the earth is not flat, but
>> this is where the projection comes in).
>>
>>>
>>> If I could get it working I could also write a plugin that could tell me
>> co-ordinates of a
>>> point on the "map" based on the new x,y co-ordinates when I re-insert
>> the microscope slide.
>>> Just in case it is not clear let me explain: When I insert a slide into
>> the EMPA machine a
>>> certain spot will have co-ordinates of 255,234,40. When I re-insert the
>> slide it might now have
>>> co-ordinates of 100,2000,80. However it is still the same point but
>> because I am measuring in
>>> nano meters small changes can make big differences.
>>
>> Just to be sure - you have three co-ordintes here? then it might be a
>> problem - but if you only
>> want to geo-reference the *picture* of the slide into a two dimensional
>> flat surface, that should
>> work easily.
>>
> 
> I do have 3 co-ordinates but for the work that we do at least for now the
> y-co-ordinate is not really important so should not be a problem.
> 
> 
>>
>>>
>>> My question is thus, does this sound possible and have anyone else done
>> something similar? The
>>> area being mapped is actually flat and I am not sure if that would make
>> a difference.
>>
>> 1) The area being mapped is flat
>> 2) are all areas parallel?
>>
>>
> Yes the area being mapped is flat for practical intents and purposes. and
> everything should be parallel. Although some images might need rotation
> since it is not possible on a micron meter scale to align up a microscope
> slide exactly parallel.
> 
> I am planning on doing a lot more work on this later in the summer and will
> report back to the list of results and any more questions which I might
> have.
> 
> Regards
> 
>