Forum

Author Topic: GCPs in the middle of nowhere  (Read 15385 times)

RalfH

  • Sr. Member
  • ****
  • Posts: 344
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #15 on: April 24, 2013, 02:10:51 PM »
frank,

thanks a lot for the link to optimalsystem. Looks like they have really good and non-expensive stuff. It's a pity their website is so poorly structured; I can't even find out whether the loggers they sell have single or dual frequency receivers.

frank.stremke

  • Full Member
  • ***
  • Posts: 206
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #16 on: April 24, 2013, 02:59:09 PM »
hi
its all single freq recievers but thats fine for rtk asl long as you have some more time :-D
hagen is a nice fellow he will be hapy to help you.
frank

RalfH

  • Sr. Member
  • ****
  • Posts: 344
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #17 on: April 24, 2013, 03:30:52 PM »
I'm not sure whether that's really fine or not. I have used a sirf-based GPS (GT11) that can record sirf binary data which can then be converted to RINEX and have post-processed these files using PPP. All of that went ok, and standard deviations reported by PPP (static mode) are usually between 0.5 an 1.5 metres, sometimes even better. But if I compare two PPP positions from the same point, they can sometimes be several metres apart. In kinematic mode, I get very many outliers and large scatter. When I asked the PPP people at NR Canada, they told me that PPP was developed for dual frequency receivers and was never intended for single frequency receivers. I think they were actually surprised that it worked at all and told me that dual frequency data are required for the correction of stratospheric effects.

For the differential post-processing with base station and rover (as, for example, with the good old Magellan Pro Mark X), single frequency should be ok, because both receivers see the same stratospheric effects. For some unknown reason, I never got the RINEX data (that were converted from sirf binary) to really work with the old Magellan differential post-processing software.

So... I'm not quite convinced how well this would work. Do you know what kind of receiver is actually used in Hagen's devices? 

frank.stremke

  • Full Member
  • ***
  • Posts: 206
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #18 on: April 24, 2013, 03:44:12 PM »
hes got different with or without glonass. i use both and it works quite good for rtk
as for the PPP with a single freq you will need of course longer observation time and corection data for post processing.
http://igscb.jpl.nasa.gov/components/prods.html
of course this way it does not work in real time but you only need to do this once for your basepoint and then use differential or RTK for the airborne system.
frank
you can also have a look here if you speek german this hforum was a big help to me
http://www.kowoma.de/gpsforum/index.php?sid=715dd77e22cf06f60b04185bcda22568
« Last Edit: April 24, 2013, 03:47:37 PM by frank.stremke »

RalfH

  • Sr. Member
  • ****
  • Posts: 344
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #19 on: April 24, 2013, 03:58:12 PM »
Thanks. I will look into all this a bit more and perhaps contact Hagen.

George

  • Full Member
  • ***
  • Posts: 193
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #20 on: April 25, 2013, 10:16:56 AM »
Hi Frank,
sounds interesting. So the method gives more precise camera positions tho. Very interesting!
What do you think about the accuracy of a Z value then?
I mean Z not the one for the aircraft altitude but for the elevation/surface i.e. the one which finally goes to a DEM.
GR

frank.stremke

  • Full Member
  • ***
  • Posts: 206
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #21 on: April 25, 2013, 12:59:22 PM »
hello,
sorry i have no answere on that question it just improves the cammera position i do not know enogh about this kind of photogrametry and photoscan. however mybe it is posible zo set down the camera position acuracy to ease up the process.
but the again there is the problem of the fast moving aircraft.
so maybe someone more wise on the internal workings of PS can help out here.
best regards
frank

George

  • Full Member
  • ***
  • Posts: 193
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #22 on: April 25, 2013, 01:16:28 PM »
actually not fast
just around 15-20 m/s

RalfH

  • Sr. Member
  • ****
  • Posts: 344
    • View Profile
Re: GCPs in the middle of nowhere
« Reply #23 on: April 25, 2013, 01:44:09 PM »
The best you can synchronise your camera with the GPS time is 1 second (I don't know of cameras that record fractions of a second in the time stamp). So if you fly at 15 m/s, the GPS position associated with the photograph may be off by 15 metres. Of course, if you fly a zig-zag pattern, you can hope that this error will cancel itself out during georeferencing. Theoretically, you could perhaps use the GPS (which can output time at millisecond precision) to trigger the camera. If the camera is set to infinity focus (focussing time is not constant, so focissing should be avoided in this case) and you can establish a constant shutter delay, you could correct for this based on the flight spped (which you can also derive from the GPS data). This will make things a bit more complicated, but if you manage to get such a setup working, it might be a solution.