Sunday, September 1, 2013

Thinking in 3D

This week at work, just for fun, I set up a stereoscope and slid some stereo photos underneath the mirrors. These photos are high resolution, large scale shots taken back in April when we had a new orthorectified aerial image of the airport developed.  I asked the contractor to send me a stereo pair from the project that I could play around with.  It's been years since I spent any time peering though the optics of a stereoscope and it was fun to look over the images and realize just how much a stereo view adds to one's ability to pick out details.



There was a time when analyzing stereo images was a critical skill in my field and other related fields.  But with the rise of commercial satellite imagery, the slow demise of wet process aerial film cameras and the development of digital imagery analysis systems like ERDAS Imagine and ESRI's improved raster management routines in ArcGIS there has been less and less call for hard copy stereo image analysis. Software routines now handle most analysis tasks.  Of course photogrammetrists still process, manage and analyze stereo imagery, but it's all done on high end digital systems these days.  The fields that used to derive benefit from hard copy stereo imagery - topography, geology, forestry, hydrology, even the US military - all seem to have lost their institutional 'feel' for the usefulness of stereo imagery analysis.

The issue was brought home to me this week when I invited a small group of GIS professionals and Engineering staff (both licensed civil engineers and engineering technicians) to drop by my desk to have a look at these stereo photos.  Most could not get the photos properly aligned underneath the stereoscope. Few recognized any real benefit from seeing the structures in stereo.  Most thought it was just a cute parlor trick. That's a shame because the stereo photos permitted quick and easy identification of features that are not readily apparent in the same 2D images.  Things like antenna masts and raised utility piping on the roofs of concourses, raised concrete pads and curbing in the aircraft gate areas and even small assemblies like receiver domes on the tops of aircraft fuselages stood out in clear detail when viewed in stereo.
.
So how does one use stereo photos for analysis?  Check out this blog posting from a while back.

Conducting stereo analysis using hard copy photos should be much cheaper and easier these days.  Years ago in the era of wet process film cameras making copies of stereo photos was time consuming and expensive. Someone had to pull a roll of film negatives, go into a darkroom and make prints one by one. With today's digital imagery systems all one has to do is download the image files from a server and print them out using relatively cheap but very high quality color ink jet printers.  The images I received from our contractor were full resolution TIFF files, each about 1.4 gigabytes.  I was able to subset just the areas I wanted to view and print them out at full resolution using only the image management software that comes with Windows 8.  Fast and cheap!

Federal and state governments are sitting on a gold mine of historical stereo aerial photos.  The Federal government (USGS, USC&GS, Soil Conservation Service, Department of Agriculture, Tennessee Valley Authority, Army Corps of Engineers, etc.) started using stereo aerial photography for mapping as early as the mid-1920s and over the course of the next 90 years proceeded to photograph virtually all of the United States in stereo.  Stereo aerial photography was the foundation of all of our topographic mapping activities through most of the 20th century and it remains so today.  Much of this photography is still held in individual agency archives or has been turned over to the National Archives. I'd love to see the National Archives digitize and post nationally significant stereo pairs of images online for downloading and viewing. Places like the Grand Canyon, Yosemite and Yellowstone or historic events like the Mount St. Helens post-eruption photos or levee breaches along the Mississippi River during the spring floods.  Even historic shots of our cities and suburbs that will help students understand how topography impacts issues like urban sprawl.

Humans view and relate to their world in three dimensions.  It's a shame that today we are relegated to investigating it via boring 2D computer displays.  I think it's time to bring back 3D image analysis!

No comments: