Scanner resolution and cell size

[Adalberto da Silva]

Hello, Rich!

I've been using only scanned map images with Grass. But I wish I can point
out some aspects.

When you scan any photo or map your maximum resolution is tied on the
*optical* resolution of course. Horizontal resolution is related to the
density of the optical sensors on the scanner trailing bar; vertical
resolution is more a function of the pass that the motor device is able to
do.
For this, you are able to acquire an image that has an square pixel, for
instance a 300 dpix 300dpi image. This means your scanner is recordind ,
let's say, only the even sensor on the bar: 0,2,4,6... and so on; and it
samples the first line, then the fourth, then the eighth. ...
For instance, my scanner has a 300 dpi x 600 dpi optical resolution (and it
works only under Windows! I'm quite sure you may be considered a lucky guy
with yours) and I captured my maps with a 300x300 resolution (square pixel),
a reasonable choice for a paper original: this means you may distinguish
between two points that are only 1/300 inches apart on the map (and I think
this is far beyond the map resolution itself).
For your photos, you must consider primarily the original resolution of the
cameras that took the photos. If they were acquired as a 1:24,000 scale,
it's reasonable to consider they can resolve 1/10 mm or 2.4 meters on the
ground. With this information you can decide the final resoltution of your
scanned picture, always taking in account the optical resolution of your
scanner.
With a 300x300 resolution you can resolve ~1/10 mm (more exactly 0.0846...
mm) or 2.4 meters on the ground; with 600x600 you increase resolution to
~4/100 mm or ~1 meter on the ground. Probably you can't get more information
than this from the photo itself as there are many constraints in the aerial
acquisition.
I hope this helps,

Adalberto da Silva

My scann

>  If this is not the proper forum for this question, please point me to the
>appropriate mail list.
>
>  My scanner has an optical resolution of 600 dpi x 1200 dpi and a plate
>size of 8.5 inches x 14 inches. When I scan standard contact prints of
>aerial photographs (9 inches x 9 inches), I capture an area 8.5 inches x 9
>inches. Now, I want to calculate the cell size at different resolutions.
>
>  First assumption: the cells are rectangular, not square.
>
>  Second assumption: the maximum horizontal resolution is 600 dpi while the
>maximum vertical resoltion is 1200 dpi.
>
>  Third assumption: I cannot make square pixels/cells directly from the
>scanner, but perhaps can in the software.
>
>  Since the standard scale of aerial photos (at least in the western US) is
>1 inch = 2000 feet (1:24,000), the horizontal extent of a scanned
photograph
>is 1700 feet x 1800 feet.
>
>  If I set scanner resolution to 500 dpi I'm setting the horizontal pixel
>density, and I will have a cell width of 4 feet (approximately 1.2 meters.)
>But, the vertical pixel density will be 1000 dpi which works out to a cell
>height of 2 feet, or approximately 0.6 meters.
>
>  Although I have scanned aerial photos before for use as underlays to the
>vector information, I've not scanned them for use as raster data (e.g.,
>vegetation classification). Are my assumptions and calculations correct?
>What do other folks do when using scanned aerial photos to define cell
size?
>
>TIA,
>
>Rich
>
>Dr. Richard B. Shepard, President
>
>                       Applied Ecosystem Services, Inc. (TM)
>              Making environmentally-responsible mining happen. (SM)
>                       --------------------------------
>            2404 SW 22nd Street | Troutdale, OR 97060-1247 | U.S.A.
> + 1 503-667-4517 (voice) | + 1 503-667-8863 (fax) |
rshepard at appl-ecosys.com
>