Hi all,
looks that with the spread of processing methods and technology,
we can make things that were available only to national security agencies, once upon a time.
Take a look at the maps of the flight made May 18th:
one day to fly, one day to upload, one day for processing, (one day for grill party), one day to download.
Few of you might remember the 3km long flight with EasyStar (a single line), stitched with Microsoft ICE.
The map above is 2.2x2.2km. It took 746 photos to make this map.
Funny fact: if the aircraft would fly straight, the surface covered would be >16km2 instead of just 4.5km2 (null side overlap).
The overlap used was 70% side, 80% along.
Comments
Thanks for the info. I'm sure the final image's relative accuracy will generally be in the 1 meter to a few decimeter range, but I was curious as to how good its absolute accuracy was without any ground control points.
This is getting more into the professional mapping side then the UAV development side, but typically you would check the accuracy by collecting some ground control points with survey quality GPS and look at the coordinate errors. Pix4d didn't seem to have much technical data on their site, but suggested that absolute accuracy might be in the 1 to 5 meter range and one example they had seemed to be a bit worse.
This becomes important when you want to overlay imagery collected at different times and do change detection analysis. Especially when you might be interested in areas of just a few, a few dozen, or even a few hundred pixels (using the great resolution you can get from a small UAV).
Krzysztof
& Lojze as well,
Thanks for sharing! Much appreciated. Great stuff!
-mikko
Well I'm quite sure that they are already working in this direction. They seem to be near that target. The main obstacle as I see is high quality automated seamline creation in urban environment where roof edge detection and detection of other obstacles is needed for high orthophoto creation.
This kind (cloud) solutions seems to popup everywhere now. Why? The computer power and GPGPU processing enable that SIFT and other methods from Structure from Motion (SfM) field produce image orientations in viable time frame. I've tried to orient 6 overlapped 216Mpx images in one of such software (Agisoft). The result of one minute processing time are oriented images and point cloud of approx. 200000 points. I know that a final evaluation should consist >100 of such images but this has to wait for now.
@John the miracle is, the final image seems to fit satellite photo within 1m just by shifting the map by hand, maybe much better. Those are multigig TIFFs and I don't have better referenced sat data at hand (we are UAV makers that try to understand every aspect of the user's work).
@Lojze impressive processing chain. You might want to contact pix4d about your processing idea, I bet they want to hear that.
Have you gone out and checked a dozen or so points in the image (sidewalk edges, fence corners, etc.) to see how good the georeferencing actually is?
I assume you aren't supplying ground control points and the software is doing a "best fit" from image GPS coordinates (autonomous) which might get you to ~5 meters horizontal error.
If so, certainly good enough for coarse mapping applications. Not survey quality, but just a few accurate ground control points - and a readjustment - could get you very close.
This is a question to http://www.pix4d.com/aboutus.html
Good job!
Could you place price for Pix4D processing somewhere?
just price range at least?
"I presume this was done with the "cloud" version of their service. Have you tested the "lite" version for UAV?"
I have tried Lite. It is:
-working, so far
-in development stage (released late March 2011 or so)
-made for lower resolution than cloud version
-made for maybe a hunderd...few hunderds protos
-made to work on your PC during the night, not on all the PCs of all your neighbors for a week
-completely automated, as with cloud (which is however assisted with very prompt user support)
-
1
-
2
-
3
of 3 Next