Hello - 
After vast amounts of work I think I have it pretty much working.  

Basic Approach:
I take the 24 scans comprising an A1 drawing and load them all.
>From the assistant I align them all - it makes a mess due to repeated 
features in the drawing.

I enable scans one at a time in the quick preview and from there for each 
one I add I optimize after modifying control points as needed. 
Eventually everything works pretty well.  I make extensive use of the 
move/drag and layout tabs, and of course of the control points tab 
in the main window.  I have expert turned on.

I use Equirectangular and Mosaic.  I crop to max so I can cut it down later 
in gimp.  I use center fit and level.

Eventually I stitch and look at the real result.

If any of this is improper practice PLEASE TELL ME.  I have about a week on 
this monster at this point hence quite ignorant.

Problems:
The size of the thing came out enormous (the physical size of the eventual 
pdf) so had to scale that back.
I still have some places where lines don't match up and I don't know why.
I have some verticals that bow on the ends which I don't understand 
either.  I have placed vert and horiz line control points on features to 
try and mitigate this.

I you are reading this trying to figure out how to make scans work on 
engineering drawings you are well advised to take the time to 
learn how to use the above tools pretty well.  They are very capable and 
about as convenient as one could get for interface - for example
in the control points editor you can easily see where you need to check a 
given scan to see if all the control points make sense.  From 
there you can also delete all those that don't make sense which is your 
only hope in getting stitch to work properly if you have repeated 
components in the drawing that confuse the align function.  

I think running align again will break everything on you and I'm quite 
unclear on whether re-optimize is the same as going to the 
quick preview window and running optimize.  I think optimize only runs on 
the active scans in that place, or rather that is what I have it 
set to and I believe that's proper.

Many thanks to all who wrote this thing and much thanks to Bruno for aiming 
me in the right direction.

Regards,
jc






On Friday, February 15, 2019 at 6:12:59 PM UTC-6, Bruno Postle wrote:
>
>
>
> On 15 February 2019 22:40:45 GMT, jim cullen wrote: 
> >Thanks for your suggestion.  After slightly modifying the 
> >sticth-scan... Python script I have an output.pto I can open in hugin 
> >and add control points.  I assume need remove points on repeated 
> >features as well, optimize and stich. 
>
> Yes, hopefully the PTO file has the optimisation parameters correctly set 
> up for a mosaic project rather than a traditional panorama - there is a 
> tutorial on the Hugin website that explains how aligning scanned images 
> works. 
>
> >and can I delete control points en masse from .pto with vim 
> >and where might I look for basic info such as this preferably in 
> >scan/mosaic (only a guess that) context? 
>
> Each control point in the PTO file is a line that starts with 'c', you can 
> delete them without corrupting the file. The file format is documented 
> here: http://hugin.sourceforge.net/docs/nona/nona.txt 
>
> -- 
> Bruno 
>

-- 
A list of frequently asked questions is available at: 
http://wiki.panotools.org/Hugin_FAQ
--- 
You received this message because you are subscribed to the Google Groups 
"hugin and other free panoramic software" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/hugin-ptx/6b581c88-e771-4868-bafe-99ec5f029ec2%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to