Jump to content

 
Photo

Scanning PDF

- - - - -

  • Please log in to reply
10 replies to this topic

#1
araki5

araki5

    Key Contributor

  • Validated Member
  • PipPipPip
  • 77 posts
  • Gender:Male
  • Interests:GIS/MAPSERVER/OSSIM/GDAL
  • United States

Hi all.
This is a problem that has come up before. I have PDF's that I have to scan and convert to tiffs. Obviously no projections or coordinate systems.

The problem I have is trying to resize and rescale when I'm in ARCMap. When you look at the size of the scanned image, it will always say in ArcCatalog, 8.5x11 or 11x17, whatever size I made it of the scan.

Obviously I need to make these pdf scans come in with a prj or tfw file so I can line them up to my highways, streets, etc.

Can anyone give a good process of doing this?

I do this all the time in AutoCad Map 3d, but Autocad Map 3d doesnt have a transparency function so I can turn off the white background. Autocad map is good at scaling and rotating and shifting, but I really need to be able to turn off the white background.

Oh, BTW, when I use the Rescale tool in Arc Toolbox, the resultant tiff is converted to an ESRI Grid Stack. This may seem weird, but when I try to rectify the resultant image, ArcMap won't recognize it anymore.

Thanks all and I hope this site stays up and running.

ARCView 9.1
Win2k
Randy Long
GIS/CAD Tech
Mackay and Somps



Raster is Faster, but Vector is Corrector.

#2
Hans van der Maarel

Hans van der Maarel

    CartoTalk Editor-in-Chief

  • Admin
  • PipPipPipPipPipPipPip
  • 3,881 posts
  • Gender:Male
  • Location:The Netherlands
  • Interests:Cartography, GIS, history, popular science, music.
  • Netherlands

So...

You have PDF files, you print them and then scan the prints to get tiff's? :blink:

What I would do is open the PDF's in Photoshop, save as TIFF, then georeference them using either Manifold or Avenza's Geographic Imager.
Hans van der Maarel - Cartotalk Editor
Red Geographics
Email: hans@redgeographics.com / Twitter: @redgeographics

#3
Martin Gamache

Martin Gamache

    Ultimate Contributor

  • Associate Admin
  • PipPipPipPipPipPip
  • 980 posts
  • Gender:Male
  • Location:Washington DC
  • Interests:History of Topographic Cartography
    Topographic Mapping
    History of Relief Depiction
    Thematic Cartography
    Demographic Cartography
    Cartographic techniques, methods, and tools
    Orienteering
    Panoramic & Kite Photography
  • United States

Hans' suggestion is probably the best for getting a TIFF file out of a PDF, open it in Photoshop and save as a Tiff.

In arcmap you can also elliminate the white background in your image by going to your image properties and changing the way the image is displayed. You can then use the image registration tools to georeference your image.

#4
araki5

araki5

    Key Contributor

  • Validated Member
  • PipPipPip
  • 77 posts
  • Gender:Male
  • Interests:GIS/MAPSERVER/OSSIM/GDAL
  • United States

So...

You have PDF files, you print them and then scan the prints to get tiff's?  :blink:

What I would do is open the PDF's in Photoshop, save as TIFF, then georeference them using either Manifold or Avenza's Geographic Imager.

<{POST_SNAPBACK}>

Hans,
What oftern happens in my line of work (Land Development, Construction etc) is that our clients get hardcopy reports from the City or the County. We won't be able to get an electronic version of the report(dumb, i know, but that's governmet for you!)
This happens for our subcontractors as well. They won't give us the file so we'll get a hard copy and scan. Kinda sucks, because we all supposed to be working on the same project.
If I do get Manifold, does it handle the scans well? I have Global Mapper, and it usually helps me out, but not in this case.

Thx Hans
Randy Long
GIS/CAD Tech
Mackay and Somps



Raster is Faster, but Vector is Corrector.

#5
Hans van der Maarel

Hans van der Maarel

    CartoTalk Editor-in-Chief

  • Admin
  • PipPipPipPipPipPipPip
  • 3,881 posts
  • Gender:Male
  • Location:The Netherlands
  • Interests:Cartography, GIS, history, popular science, music.
  • Netherlands

What oftern happens in my line of work (Land Development, Construction etc) is that our clients get hardcopy reports from the City or the County. We won't be able to get an electronic version of the report(dumb, i know, but that's governmet for you!)
This happens for our subcontractors as well. They won't give us the file so we'll get a hard copy and scan. Kinda sucks, because we all supposed to be working on the same project.


For those situations, I recommend the pointy stick method... Or staging a Mission Impossible-style raid on their offices... :ph34r: (yes, finally an opportunity to use that one!!!)

If I do get Manifold, does it handle the scans well? I have Global Mapper, and it usually helps me out, but not in this case.

<{POST_SNAPBACK}>


It handles raster images quite well and, if you have a corresponding georeferenced vector file, it's fairly easy to georeference a raster. Just pick some control points and specify the projection.
Hans van der Maarel - Cartotalk Editor
Red Geographics
Email: hans@redgeographics.com / Twitter: @redgeographics

#6
Martin Gamache

Martin Gamache

    Ultimate Contributor

  • Associate Admin
  • PipPipPipPipPipPip
  • 980 posts
  • Gender:Male
  • Location:Washington DC
  • Interests:History of Topographic Cartography
    Topographic Mapping
    History of Relief Depiction
    Thematic Cartography
    Demographic Cartography
    Cartographic techniques, methods, and tools
    Orienteering
    Panoramic &amp; Kite Photography
  • United States

It sounds to me like you already have all the tools you need to get the job done in terms of georeferencing the scanned images (arcmap or global mapper) and that Manifold will not provide you with anything new.

#7
araki5

araki5

    Key Contributor

  • Validated Member
  • PipPipPip
  • 77 posts
  • Gender:Male
  • Interests:GIS/MAPSERVER/OSSIM/GDAL
  • United States

GM let me get it - after about 4 hrs!

This really takes more skill than i thougt! But I DID notice that when you choose only 2 pts as opposed to 4 or more, then the warping is hardly noticable.

Thanks for the tips guys!
Randy Long
GIS/CAD Tech
Mackay and Somps



Raster is Faster, but Vector is Corrector.

#8
Hans van der Maarel

Hans van der Maarel

    CartoTalk Editor-in-Chief

  • Admin
  • PipPipPipPipPipPipPip
  • 3,881 posts
  • Gender:Male
  • Location:The Netherlands
  • Interests:Cartography, GIS, history, popular science, music.
  • Netherlands

GM let me get it - after about 4 hrs!

This really takes more skill than i thougt! But I DID notice that when you choose only 2 pts as opposed to 4 or more, then the warping is hardly noticable.

Thanks for the tips guys!

<{POST_SNAPBACK}>


It's more reliable to pick at least 3 points though.
Hans van der Maarel - Cartotalk Editor
Red Geographics
Email: hans@redgeographics.com / Twitter: @redgeographics

#9
Kartograph

Kartograph

    Legendary Contributor

  • Validated Member
  • PipPipPipPipPip
  • 320 posts
  • Gender:Male
  • Location:Berlin, Deutschland
  • Germany

It's more reliable to pick at least 3 points though.


Well, that really depends. If it?s an affine transformation technically 1 point and the dimensions of the picture would just be just as accurate.

Could you elaborate on the GM problems you had? I?ve had very good experiences with their georeferencing tool so far.

#10
araki5

araki5

    Key Contributor

  • Validated Member
  • PipPipPip
  • 77 posts
  • Gender:Male
  • Interests:GIS/MAPSERVER/OSSIM/GDAL
  • United States

It's more reliable to pick at least 3 points though.


Well, that really depends. If it?s an affine transformation technically 1 point and the dimensions of the picture would just be just as accurate.

Could you elaborate on the GM problems you had? I?ve had very good experiences with their georeferencing tool so far.

<{POST_SNAPBACK}>

When I originally tried to georeference the scanned image, I used 4 pts(that is the default that is in the dialog box when you want to georeference), the picture was really warped, like not recognizable at all.

So I ended up picking 2 pts, one in the far northeast corner that I could positivey get coordinates on and I took another point in the farthest southwest corner that I could get coordinates on.

It came in really good for the 2 point approach.

One thing that is still bugging me, is when you scan and you use photoshop(i used photoshop 6.0) Arccatalog keeps the original dimensions of the scanned image, this case 11x17 cropped down to 8x8 so as to make a square.
But when I went to GM and did the georectify, it's like it automatically takes the image dimensions and will "include" the dimensions for use in arcmap or autocad or whatever. Really neat how it does that. When you look at the metadata in ArcCatalog, the original scan the "projected/local coordinates 11.00in right, 17.00in left." Now the "georectified" scan has actuall coordiantes, in this case CA NAD83 State Plane zone II, 6,200,000 and 2,100,000 (roughly).
It really is pretty lame that ESRI georectifying abilities are WAY under that of GM (and from what I'm told, Manifold).

Thanks for the help
Randy Long
GIS/CAD Tech
Mackay and Somps



Raster is Faster, but Vector is Corrector.

#11
Martin Gamache

Martin Gamache

    Ultimate Contributor

  • Associate Admin
  • PipPipPipPipPipPip
  • 980 posts
  • Gender:Male
  • Location:Washington DC
  • Interests:History of Topographic Cartography
    Topographic Mapping
    History of Relief Depiction
    Thematic Cartography
    Demographic Cartography
    Cartographic techniques, methods, and tools
    Orienteering
    Panoramic &amp; Kite Photography
  • United States

...One thing that is still bugging me, is when you scan and you use photoshop(i used photoshop 6.0) Arccatalog keeps the original dimensions of the scanned image...



Can you describe how you are georeferencing in ARC.

I've never had the problem you describe and I find Arc's georeferencing tools equal to Manifold's ( I have not used GM, but if it takes 4hrs to do one sheet it is not for me) and neither take more than 10 minutes to process an image. I get the feeling you are missing a step somewhere. When you refer to the size in arc catalog is this for the image after you have used the georeference tools and saved a georeferenced image?

Regarding the number of points; A first order affine transformation usually requires 3 points in most applications that do this, but greatly benefits from more, well distributed points. The difference in processing between 2 and 4 points should not be very great for an affine transformation.
Warping will only occur in 2nd and 3rd order polynomial transformations (which require alot of points to give good results). In a first order transformation all you are doing is scaling and rotating so there should be no torsional distortion.

Arc and other software reports a RMSE which can provide some feedback as to the quality of the points.


mg




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

-->