Hugo,
I'm keen to assist (and curious to see what will come out of it).
By the way: I have some local (Johannesburg) friends working on a similar
UAV project ... maybe there is space for synergies and/or exchange of ideas?
From our side: when we managed to gain access to some 6Tb of raw 36bit
aerial data (and metadata) from a professional digital camera we studied the
subject in detail and eventually developed a TileService module which in
automatic would use metadata flight information for georeferencing
individual aerial photos and then othorectify (using a DTM), colour balance
(using the flight metadata) and mosaic them together into tiles ready for
web serving.
You are welcome to examine the output of one such test:
http://warsaw.freeearthfoundation.com/nowak/testmap/?x=0.551184207201004&y=0.34431183338165283&sx=8388608&sy=8388608&layers=bmng_topo_bathy_200401,test.capeortho2,&layer_manager=0
By looking to the details you can possibly have a better idea of the various
issues at play.
In the above example the morphology was extreme: from sea level to 1050m
above sea level within a few hundred meters. To our amazement we found no
use whatsoever for any ground control point (they just .... complied). That
should not have come as a surprise, after all aerial photography is a
science not a phylosophy: if flight metadata is good (and the processing
maths correct) ground control points become irrelevant and are simply for QA
purposes.
FLIGHT METADATA IS THE CRITICAL ISSUE: your final results will be as good as
your flight metadata is.
Don't ask: I have no idea how that data is actually collected :D
I can guess however that as a fair amount of googgling was critical in us
understanding how to use that flight metadata so a fair amount of googling
should help you in understanding what kind of data to record and how.
It might sound really complicated and even scary: you will find that the
main stumbling block is the initial fear for the subject rather than the
implementation of the concepts in practice.
Ciao
Maurizio
----- Original Message -----
From: "Hugo Meiland" <h...@meiland.nl>
To: "talk" <talk@openaerialmap.org>
Sent: Thursday, December 10, 2009 12:02 AM
Subject: [OAM-talk] starting to map with small UAV
Hi all,
I've been playing around for a long time with model airplanes and both
openstreetmap and openaerialmap got me interested into combining this
hobby with some fun web stuff.
What I would like to do is to make a detailed set of aerial pictures and a
map available of a nearby strip of dunes. For this I have made some
'manual' flights with a strapped-on digital camera to see what the results
could be. Now I am building a larger plane which should be flying with an
autopilot to allow more controlled height and straight lines.
As far as I know the following steps I should be taking is to:
- combine the images into a larger image and
- then to orthorectify that image.
- once that is done, geodata should be attached to the image, and
- I should be able to store this in a WMS
- and create tiles from the to serve to e.g. openlayers
Am I correct in the above workflow, or missing steps?
thanks!
Hugo
_______________________________________________
talk mailing list
talk@openaerialmap.org
http://openaerialmap.org/mailman/listinfo/talk_openaerialmap.org
_______________________________________________
talk mailing list
talk@openaerialmap.org
http://openaerialmap.org/mailman/listinfo/talk_openaerialmap.org
