Updated for the Mapping Workshop at Cleveland Public Library, November 18, 2009
Thomas Edwards, Map Librarian, Cleveland Public Library
Stephen Titchenal, Retired Technology Teacher/Librarian
Scanning pixels (dots) per inch refers to the number of light samples (pixels) taken per inch of the original. Print resolution determines the size of the image on paper which could be different than the original scanning resolution. Computer monitors can display about 72 to 100 pixels per inch. Scanning at higher pixels per inch will preserve more details when viewing or printing enlarged areas of the original scanned image. 300 dpi should be the minimum scanning resolution considered. Photos or drawings with tiny details should be scanned at higher resolutions (600 - 1200 dpi -- even higher for slides and negatives)
TIFF is the standard for archival digital images. The idea is to make a high
resolution scan that can be used to create images in other formats for specific
purposes. The original maps never need to be scanned again. LZW compression
is lossless, if file size is an issue. For archival purposes, never use a lossy compression type.
Because low end scanners are not perfect and storage space can be an issue
for large map images, you may decide to use a lossy file format. PDF and JPG
are the two common formats that allow you to adjust the size vs quality of the image. I suggest scanning at a higher
resolution to offset the increase in unwanted artifacts and loss of detail possible with these formats. Each time
you save an edited version in these file formats the image is noticeably changed.
If you intend to edit and save an image multiple times you should temporarily
save in tiff format. Always preserve the original scan as a backup.
I normally scan even grayscale/b&w maps and drawings in color. This makes
it easier to remove stains, marks and blemishes while preserving the original
inks. If an original b&w image does not need cleanup, then grayscale will
save space and still allow threshold adjustments later.
Specialized display formats: DjVu, MrSID, JPEG2000, PDF
These formats can perform much higher lossy compression, while still preserving
details. They require a separately installed viewer plug-in for display in
a web browser. For the end users, DjVu viewer seems to display large maps much faster
than the other formats. MrSID is often used for aerial images. JPEG2000 is
used by ContentDM as the master image to render on the fly for display in a
web browser without need for a separate viewer. PDF is the most common add-on format for files on the web but has trouble with larger files. Download larger files to your hard drive and open directly in Adobe Reader/Acrobat rather than displaying in the browser.
1. Burn multiple DVDs of original and edited masters.
2. Copy to multiple hard drives. External USB or eSata hard drives are very inexpensive.(~$100 a terabyte)
3. Re-copy files every few years - comparing file size after copying.
Creating image overlays for free Internet Mapping services
Each Internet mapping service or GIS (Geographic Information System) uses a specific projection to display the three dimensional spherical Earth on a flat surface. When you match points on the scanned map to known points on an image or drawing already projected in the GIS, you can overlay the scanned image. This is called georeferencing or rectifying the scanned map or aerial photo. Each of the major Internet Mapping services have a free utility to georeference scanned images
Microsoft Bing (Virtual Earth): Map Cruncher tips (http://research.microsoft.com/mapcruncher/)
Advantages: free, easy to use, generates multiple zoom levels to speed display
Disadvantages: multiple zoom levels can increase total size and rendering times.
Must have space on a file/web server to share over network.
The help file that comes with map cruncher is very well done. These tips highlight
choices that are easy to overlook the first time you create a map set.
higher resolution the original map image is, the sharper the map will be at
higher zoom levels. But the time to render and the number of tiles
will increase exponentially. When rendering choose a zoom level that complements
the resolution of the map. Zoom level 19 corresponds to street level. Only
zoom to that level if your map was drawn and scanned with that level of detail.
2. A layer is a single tile set that is always viewed at once. It may comprise
a single source map, or multiple source maps. If you want to be able to turn
individual maps on and off put them in separate layers.
Google Earth/Maps tips (http://earth.google.com/)
Advantages: Google Earth is the most interactive of the free Internet based mapping tools. Offers many advanced options with
KML. Version 5.0 includes timeline with multiple historic aerial photos (currently from 1990 to the present in some areas.) Ability to turn on/off or fade multiple layers. Images georeferenced in Google Earth can also be viewed in Google Maps.
Disadvantages: Higher learning curve for some features. Rural areas do not
have as high a resolution satellite images as Microsoft Virtual Earth.
Tips: The graphics card in the viewer's computer determines the largest image size that can be displayed. Older computers may have trouble with images larger that 2048 pixels in length or width. Newer computers can support 4096 pixels. The larger the image size the longer it will take to load. Superoverlays refers breaking an image into smaller tiles at each earth display resolution. This can be done using Global Mapper or other commercial products.
Image Overlay Tutorial http://earth.google.com/userguide/v5/ug_imageoverlays.html
Simple overlays can be viewed in Google Maps by pasting the URL of the KML/KMZ file into the Google Maps search box. Rotation is not supported. By choosing "Save to My Maps" they can be saved for future use. Importing overlays directly into My Maps does not appear to be supported yet. The new iPhone Google Earth app currently supports user created My Maps but not those created by others.