Angela Sanders, project coordinator, National Geographic Research, Conservation, and Exploration Group
The National Geographic Africa Aerial Survey project (Megaflyover) was a joint effort between National Geographic and the Wildlife Conservation Society (WCS). The project's goal was to "compile a visual record of Africa's environment, from the most densely packed places to the wildest," to validate assumptions of a larger initiative, the Human Footprint project. This larger project assesses the impact that humans are having on the planet.
The results of the Megaflyover project were published in a special single subject issue of National Geographic, September 2005, entitled "Africa: Whatever You Thought, Think Again." National Geographic reported: "Through its Human Footprint project, the Wildlife Conservation Society has created a global map of human influence on the earth's surface. WCS staff used satellite imagery and census data to map population, land use, roads, lights at night, and other variables." For more information, please visit the National Geographic Africa Megaflyover page.
Dr. Michael Fay, explorer in residence at National Geographic and conservationist for the WCS, conducted the seven-month flyover across the continent of Africa, covering 60,000 miles and 21 countries from South Africa to Portugal. A Cessna 182 airplane flew low—about 500 feet from the ground—shooting with a high-resolution Kodak DCS Pro 14n DSLR digital camera connected to a GPS device. A photograph was automatically taken every 20 seconds with location information embedded in each image.
Recently, we spoke with Angela Sanders, whose job was to organize, edit, and select the final images from the over 100,000 aerial photos captured on this project. She described the process from the beginning:
"You can imagine the complicated tangle of wires in (Mike's) cockpit. Photos were uploaded into an HP [Hewlett Packard] tablet computer on his lap, through which he could add notes and annotate his observations below. Mike was viewing images and annotating them as they were captured. He was able to caption approximately one out of every ten images. He observed, for example, that the livestock below him in some cases had crossed over to protected areas; he was able to note the extensive land erosion below. The computer was backed up to external hard drives nightly.
"I came onto the project several months after Mike had returned and had transferred and consolidated all the images onto two one-terabyte external hard drive disks, each containing about 50,000 images. To start, we created two iView MediaPro catalogs, one for each hard drive. The filenames had been renamed using the EXIF capture date (year, month, date, hour, minute, second) allowing for a unique name, such as 2004-08-17-09-46-54.DCR (Kodak Raw file).
"Then I began the sorting process. The first step was to correlate images with a country. Using Mike's GPS flight log data, we could determine the exact time and position the plane crossed international boundaries, and in turn we could correlate that time to the timestamp and file name of the images.
"Much of what I did was to focus on image quality. Over the following weeks and months I began a cataloguing process within each of 2 iView Media Pro projects. The following steps outline my workflow for the second .ivc file, which contained a set of 35,000 images:
- I created custom catalog sets for ground, lateral, and vertical shots (I worked with vertical sets first and most because these were the geo-referenced images that provided us with the "postage stamp" view of the earth—the visual record to validate the Human Footprint model). The vertical catalog set comprised about 31,000 images.
- From these, I created a catalogue set for unusable images (images that were damaged, blurred, pavement shots from takeoff and landing … a total about 250).
- Then I created another catalogue set for water images. For example, the flight across the Mozambique Channel yielded several hundred featureless blue images.
- This left me with a set of about 30,000 images, which I put in another catalogue set: the fourth select useable.
- From these 30,000 images, I created another set: the third select representative (around 10,000 images). This set was comprised of images with interesting features, anything beautiful, anything Mike annotated, and at least one image from every series of seven or so to ensure that this select was representative of the entire set.
- I created a set for the second select good composition (around 5,000 images).
- I created a set for selects (2,200 images).
- I created a set for primo selects (900 images).
"Part of the unique nature of this project was the astounding resolution of this huge set of images, significantly better than the naked eye could see from the plane.
"Our partner, WCS, used geographic information systems (GIS) software to analyze the photographs and create another database. The position data embedded in the images provided 20 new pieces of data to attach to each image. These data included fields such as the ecoregion in which the photograph was taken and if a protected area was associated with the images. [Note: the World Wildlife Fund is credited for publishing research on the world's ecoregions. The term ecoregion comes from this research.] Most importantly, these data also showed what the Human Footprint value was for the location of each photo. This information was essential for project because the goal was to verify the location of the wildest places as defined by the Human Footprint map. The images would either verify or challenge the map's assumptions.
"The result of both organizations' work was two separate databases to correlate—our image database and the WCS analytical database. Initially, I had to go back and forth searching between the two different databases to utilize all of the information available for the photographs. For example, The Green Belt Movement in Kenya contacted me in search of images from our project captured in the region of the Aberdare Mountains. To locate this specific group, I had to correlate between the WCS database with the geographic information and my iView catalogue to identify the images."
"iView Multimedia worked with us and created a custom metadata-management script that could import text files, create custom fields and place the data in the custom field very efficiently. It allowed me to import and correlate all the data I had from two different organizations and multiple different spreadsheets into one place. Now I can keep all of the information in one MediaPro catalog.
"This project clearly called for the cataloging abilities of MediaPro. But beyond this project, I find that MediaPro is useful for a wide variety of day-to-day tasks. For example, we regularly use it to track what images have been submitted to which collections. MediaPro allows us to quickly and easily caption images, use keywords, and sync metadata—all powerful organizational tools, without which our work would be incredibly tedious and unacceptably time consuming. It is the key to our archiving process because MediaPro permits us to quickly and easily access, organize, repurpose—that's to say, make the most of—our digital media assets."
