Forum

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - adamski

Pages: [1]
1
General / Re: OT: GPS guide for aerial mapping in full scale aircraft
« on: April 03, 2012, 04:30:19 PM »
Dino - do your flight lines really need to be followed very accurately? Or are your pilots good enough to follow a path to within metres?

The following applies to a full sized helicopter system:

We've used a standard mapping GPS to set up flight paths and have them followed, just using waypoint-to-waypoint track following. This requires a lot of pilot training and a very good relationship with the navigator/instrument operator. It is also very hard work for a pilot to keep a helicopter flying straight and level - although making their navigation screen bigger (ie an ipad/rugged tablet) and using some moving map software (eg oziexplorer, or something like: http://itunes.apple.com/au/app/igmap-fswidgets-moving-map/id379872511?mt=8)might help.

But - we don't worry about following flight lines super accurately because the pilots can't physically do it. To compensate all the data we collect are geolocated using an onboard dual-frequency GPS+IMU. If we're flying grids we aim to also collect redundant data to further offset the fact that flight lines are not very accurate (metres), and cannot be followed very accurately anyway (metres to tens of metres).

So if you're relying on the pilot following a line I think you're in trouble, your apriori camera positions will potentially be way off. Quick and dirty solution might be to use something like a Nikon DSLR and its GPS extra bits, so you can have each image tagged with a (good but not great) GPS timestamp and position.

Adding ground control probably goes a long some way to fixing this. My (small) experience so far with photoscan is that it is best to give the images decent apriori camera locations, then use ground control to tighten everything up a bit. However, my datasets generally do not have much, if any, ground control. I'm relying very much on the camera centres and a couple of known objects at one end of a flight.

Hope that helps!

2
Feature Requests / Re: Command line access
« on: March 05, 2012, 01:58:26 AM »
This looks to be an oldish request, but please consider this an additional 'yes please, add command line support'.

As per WickedAndy, command line access with options would be great.

I'd also be really happy to the Python interface, for example laying out all the required options in a python script and typing 'runMyPhotoscanJob.py' into the terminal.

Thanks!
Adam


3
Python and Java API / Re: starting
« on: March 05, 2012, 01:54:23 AM »
Hello Jan,

Currently PhotoScan Python module is only accessible from PhotoScan GUI.

Ah! I am in the same position as Jan - I would like to run photoscan on a big compute box from a remote machine so invoking the whole process from a command-line script is very desirable!

I take it, that this means I need to open the photoscan GUI over the network in order to make it go?

Thanks

Adam

Pages: [1]