Forum

Author Topic: Digitisation workflow for historical aerial photographs...  (Read 7463 times)

bigben

  • Sr. Member
  • ****
  • Posts: 406
    • View Profile
Hi All

I'm doing some tests for scoping a large digitisation project of historical aerial photographs (20,000 6"x6" negatives) and figured I should include steps in the workflow to prepare them for use in Photoscan or similar applications.

Digitised at 16bit greyscale
Levels adjustment (image of clock adjusted separately as it's quite dark)
Align images to template (we'll be using PTGui for this. Only offset and rotation will be optimised)
Output aligned images at fixed pixel dimensions
Keep at 16bit or convert to 8bit?

Anything else I should consider?

Regarding masking in Photoscan, I've done a reconstruction using only the central portion of the image.  Is there any significant value in including the side and bottom sections as well?  I've only got 4 negatives from a single flight path so I don't have much scope for testing. I can probably get PTGui to calculate a correction for falloff of the lens as well.

And while not related to Photoscan, any recommendations on scanning equipment or services that might do this with specific relevance to photogrammetry?  The collection is in Georgia

Quick test : https://skfb.ly/DNUp
« Last Edit: May 05, 2015, 03:21:11 AM by bigben »

yelnam

  • Newbie
  • *
  • Posts: 2
    • View Profile
Re: Digitisation workflow for historical aerial photographs...
« Reply #1 on: May 20, 2015, 09:28:28 PM »
From what I've heard, it's terribly important not to crop or otherwise modify the scans.  You'll need the entire photo, without subsequent manipulation, to have PS align the photos with automatic triangulation ... .

aggieair

  • Jr. Member
  • **
  • Posts: 93
    • View Profile
Re: Digitisation workflow for historical aerial photographs...
« Reply #2 on: May 20, 2015, 10:54:44 PM »
    This summer I will embark on a very similar project with a LOT of photos. haven't figured out my plan of action yet.

    I did get some scanner recommendations from the USDA-FSA-Aerial Photography Field Office as far as using a large format reflective scanner, from his email:

Hard copy photos will need a reflective scanner.  Most flatbed scanners are about 8.5” x 12”.  Since your aerial photos are approximately 9.5” x 9.5” you will need to determine if a standard scanner would cut off too much of the photo or if you could consider a scanner designed for larger documents.  These are often 11” x 17” or larger.  Like traditional scanners the price can vary drastically.  Flatbed scanner designed for larger documents.  Some examples can be seen at:
http://www.amazon.com/s?ie=UTF8&page=1&rh=i%3Aaps%2Ck%3Alarge%20document%20scanner
 
My recommendation would be that you investigate:

++ Traditional 8.5 x 11 scanners (may not get the edges but aerial photography usually has 30 – 70% overlap)
++ Large document scanner 
++ Take photographs of the photos with a DSLR camera and use these as your scanned image.  This could work well, especially if you build a frame to keep the angle and distance consistent.[/li][/list]

 
Remember that paper photos may or may not have the PPI (points per inch) that would enable a 15 micron (1693 dpi) scan.  Film will almost always have a much higher level of resolving capability than a paper photograph.

 
 
 


bigben

  • Sr. Member
  • ****
  • Posts: 406
    • View Profile
Re: Digitisation workflow for historical aerial photographs...
« Reply #3 on: May 21, 2015, 03:23:39 AM »
The only modification I'm planning is to align the images to the markers on the film.  Imaging with a camera does throw in an added layer of complexity as the image has then been produced by two lenses. To make things a bit more consistent I'm looking at centering the negatives in the camera frame and using a plastic overlay on the monitor to mark out the positioning of the negative in the live view of the camera interface. That will remove the need for subsequent offset and rotation adjustment and hopefully minimise the impact of imaging through two lenses.

As for scanner selection.  There is another important factor to consider which is scanning time.  We have an Epson 10000XL which does a nice job but a 1200 dpi scan of a 6" negative takes 5 1/2 minutes.  That's nearly 80 days of scanning time alone. Taking into account sleeping, eating, breaks and life in general that drags out to 320 days of just scanning.  On the basis of that, we ended up purchasing a Hasselblad H5D60 for digitising film larger than 6cm (we have other projects on the way that aren't aerial photography)  This is not going to get the same resolution as the Epson for 6" negatives, but I'll be doing a small project with the sample we have with a 35mm DSLR, the H5D60 and Epson scanner to compare workflow efficiency and outputs.

To go up to 1693dpi while keeping native resolution steps in the scanner would require 2400dpi, but that would be unrealistic for 20,000 unless you had a lot of them.

Another consideration with the flatbed scanner is that the negatives curl a fair bit, so we would have to devise a way of keeping them flat(ter) which then adds to the handling time.  I've just received a batch of sample images with known locations. These were scanned on an Epson going by the metadata, and exhibit the same curling which is obvious in the images.

StevenF

  • Newbie
  • *
  • Posts: 39
    • View Profile
Re: Digitisation workflow for historical aerial photographs...
« Reply #4 on: May 22, 2015, 02:28:50 AM »
Hi Bigben,
You're using PTGui to align the fiducials of images taken with the same camera and lens, correct? Could you could provide a brief overview of your workflow for this step? I haven't used PTGui before, but I might try to implement a similar step in my own work with historical images.

Since you're working with non-photogrammetric scanners/cameras you might want to consider calibration of the scanner so you can remove it's distortions before correcting for the camera lens distortion in PhotoScan. It might also save you some time to develop an auto advance mechanism if you're working with rolls of film. Someone I'm working with is implementing these steps into their workflow for processing historical imagery.

bigben

  • Sr. Member
  • ****
  • Posts: 406
    • View Profile
Re: Digitisation workflow for historical aerial photographs...
« Reply #5 on: May 22, 2015, 05:45:49 AM »
Hi Steven

Correcting the lens distortion of the digitisation camera makes sense.

For the alignment, I load a template image which is just a greyscale image with the fiducials. It doesn't need to be the same size as the scans, but it does make life easier as the zoom controls are not independent in the control point screen.  Then add all of the scans and select control points for the fiducials between the template image and each scan.

In my case that's 4 control points per image pair which is not enough to do the optimisation in one step, so I run separate optimisations for HShift,  VShift and Roll.

Output individual layers only.

The flatbed scans I got were all cropped to different sizes so I also optimised FOV for each image.  As the scans are optically different from a camera image I just specify a very small fov for the images (1°) so that the differences in projections are minimal.  You could go even smaller if you wanted. PTGui also has optional interpolators for higher quality rotation.

You then end up with all of the fiducials in the same point in each image, so it's quite easy to create a single mask image that will mask out any extraneous image areas.

But selecting control points on large images is a bit of a painful process which is why I'm leaning towards the screen overlay and aligning the images via live view at the time of capture.  We'll see which is more annoying to do over and over.

..... Wouldn't it be nice it PS had a little magnifier window like PTGui for adding GCPs. Saves zooming in and out for precise placement