Comments

• 

  Christian Stallings May 5, 2014 at 2:54pm

  I would recommend looking at the bookstore of asprs.org. They have several books that would be good resources to understand basic photogrammetry. As for the software, photoscan is a good program to start with. It’s not as robust as most UAV suites but it just works. The leap from a suite like this to one we use for professional photogrammetry like Intergraph ISAT or Trimble Match-AT is going to be about $30,000 and they can be a lot more complicated and ultimately have more tools than you will ever need for a UAV.  Hope this helps.
• Jason Best May 5, 2014 at 10:59am

  Christian, thanks for your response. I'm quite new at photogrammetry so some of the solutions you mention are not familiar to me. Is there an introductory text or guide that you could recommend for someone like me to get up to speed on the application of these techniques? The only way I've been able to get as far as I have (generating 3d models and orthophotos with VisualSFM) is thanks to various step by step tutorials and guides so I don't have a deep background in this to improvise something I haven't seen before. Thanks for helping a newb!
• Troy Gere May 1, 2014 at 2:56pm

  That will give you the x,y,z values of the camera's physical location at the time the image was captured.

  This is referred to as the "Photo Center"
• 

  Giovanni Gadaleta May 1, 2014 at 2:52pm

  I have a question for you: if you use a Canon SX260 could you tell me the GPS coordinate at which point of the picture refers?

  Thank you very much.

  Giovanni
• 

  Christian Stallings May 1, 2014 at 7:09am

  Traditionally in photogrammetry this is how we solved AT. First you need to solve for your interior orientation and tie your imagery together. Then using ground control you can create a block adjustment. This same process pretty much holds true with modern UAV image processing suites. I know that one approach you can take is to use photoscan to solve for your camera and tie your images together. It does not need starting GPS points to tie images together. Once you have this you can add ground control. This can either be true ground control that was collected or you could use known points from google earth. Obviously the second choice will not be very accurate at all however if you are just looking for rough orientation, this should work. Once you apply those ground control points, photoscan should create an exterior orientation that can then be used in other software. I hope this helps.
• Jason Best April 30, 2014 at 3:44pm

  Does anyone have any experience, solutions, or suggestions for determining GPS coordinates of images that have been used to create an orthorectified image generated using SFM? In more detail here is what I'd like to do: I've performed a number of aerial surveys using kites and balloons and a camera, but I don't have GPS (or orientation) data for each image. I then took those survey images and stitched an orthoimage using VisualSFM, then orthorectified that in ArcGIS. So I want to essentially take the GPS coordinates I now have for the stitched image and take that back through some process to determine the coordinates of each original image. I'm assuming there would be a way to do this using the VRML model that VisualSFM can create or maybe through the various intermediate files like the point cloud files. Ultimately the reason I want to do this is to be able to reprocess these individual survey images in software like Pix4UAV which (as far as I can tell) requires GPS coordinates for each image. Any thoughts are appreciated!

  - Jason
• 

  Marek June 6, 2013 at 11:44am

  in f2.zip (134 MB) - 60 photos, Canon SX230HS, altitude 100m, 4000x3000 px, area 0,25 km2
• 

  Marek June 6, 2013 at 11:40am

  here is link for sample photos with GPS data in EXIF:

  https://www.dropbox.com/s/dtqjuqlli10w4p8/f2.zip

  

  Dropbox - Link not found

  Dropbox is a free service that lets you bring your photos, docs, and videos anywhere and share them easily. Never email yourself a file again!
• 

  Drone Nerd June 6, 2013 at 9:57am

  Christian,

  Just PM me, and I send a link, about 270MB, 300 shoots, is worth?
• 

  Christian Stallings June 6, 2013 at 9:44am

  Hi,

   I'm looking for some image datasets that I can use to test various auto correlation techniques. If anyone has any they are willing to share, I would be truly grateful. Specifically, if anyone has data sets of any type of right of way, that would be extremely helpful. Finally a dataset with nadir, forward and aft obliques would be awesome. Any help is greatly appreciated.

   

  Thanks