Wednesday, November 13, 2013

[CRYOLIST] More aerial imaging and DEMs - this time on sea ice (mostly)

Hi Cryolist

Following Matt Nolan's recent post (great read, and great work, by the way)..

We are using the exact same technology for some sea ice work. Here's a small summary:


I wanted to add to Matt's comments a little too.

Firstly, if a GPS-enabled camera is not available, this technology does an excellent job of georeferencing (to reasonable accuracies) the end products using a very sparse set of ground control. I've done some testing over an Australian Antarctic station, to test the usefulness of a totally ad-hoc aerial photo capture mission (think - its a good day, lets fly and get a pile of pictures, but don't have a whizzbang GPS camera...) - with reasonably good results. 

Of course, there are many things to consider - accuracy in model scales, requirements for matching other survey data, and so on. Since models can be scaled independent of its location (eg, using a known object length in an image), the balance between absolute georeferencing accuracy and internal scale accuracy (ie things in the model are the right size) can be tweaked pretty well.

Next, we've found that flying single strips is not a good idea, even with a lot of image overlap - so a typical sea ice flight campaign is not an ideal scenario. A multi-view camera system might help - but you really want to cover an area well, and from different angles, to get good results.

Cheers
Adam

--
Adam Steer

PhD candidate :: Data analyst
School of Geography and Environmental Studies :: Antarctic Climate and Ecosystems CRC
University of Tasmania
http://www.utas.edu.au/ :: http://www.acecrc.org.au

at