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 - WickedShell

Pages: [1] 2
Bug Reports / Re: Unable to launch on a Windows 7 machine.
« on: March 02, 2016, 10:41:32 AM »
I do have Kaspersky installed there :) But I also have it installed on another computer that isn't having any problem. Nothings jumping out as me incorrect on that machine, I granted Agisoft unlimited network access on that machine and its still not working right. Do you know what specific part of Kaspersky might be causing interference?

Thanks for you're help so far.

Bug Reports / Unable to launch on a Windows 7 machine.
« on: March 01, 2016, 01:06:36 AM »
Recently I have on windows 7 machine which is completely unable to launch Agisoft, when I launch it I get a empty cmd prompt with no text that hangs forever but does have the correct icon set. This has been happening with 1.2.3 although was happening before I upgraded versions trying to fix it. I've reinstalled it several times and that hasn't seemed to help. This computer did run agisoft without any significant problems for some time, before it stopped booting.

Let me know what additional info is needed to debug.

Windows 7 Ultimate SP1
Xeon E5-1620 v3 @3.50GHz
Nvidia GTX 980 4GB

Python and Java API / Re: Install python libraries
« on: May 25, 2014, 05:38:53 AM »
For python libraries that heavily leverage is there any way to run them in that configuration, or  simply try and force it to work by moving in the folder from site-packages?

Thanks for your time.

Python and Java API / Install python libraries
« on: May 15, 2014, 04:41:15 AM »
I'd like to install several python libraries on top of the standard agisoft libraries, how would I go about doing this? (The libraries would typically be installed with

Python and Java API / Re: matchPhotos() not using all CPU cores?
« on: May 07, 2014, 06:22:23 PM »
I'm suprised the memory consumption would be that much higher then some of the later steps (and the final step of detecting points), but obviously I'm unfamiliar with that side of the code. On the current machine for 2587 images it took 1 hour 3 minutes to detect points, and 1 hour and 51 minutes to select pairs and match (I believe it was one hour and 51 minutes respectively). Which is the only reason I'm looking to speed that up.

The workflow has moved into reconstructing depth at the moment, (no GPU attatched), and I'm only seeing 75-80% CPU usage, is this the same behavior? (and if so this would be really nice to be able to modify it here, as I'm looking at 21 hours, 40 minutes to build depth, when it could be leveraging 20% more CPU.

Python and Java API / matchPhotos() not using all CPU cores?
« on: May 07, 2014, 10:12:02 AM »
I'm using a machine that has 16 true cores, and due to hyperthreading Linux enumerates it as 32 cores. Agisoft is only appears to be using 8 of the 32 enumerated cores. (Based on watching top). Is there a limit to the number of CPU cores it can actually utilize or am seeing lower CPU usage based on limiting factors from the disk?

<speculation> Actually, based on watching the console it is strictly handling detecting points on a per photo basis in a strictly monotonic order, which leads me to believe that agisoft is throwing all the cores at a single image at a time, and then waiting to sync the results before starting the next one. If this is the case then could would it be possible to instead handle point detection on an image per thread basis? (A lot of CPU time is sitting idle at the moment. </speculation>

Python and Java API / Re: Running headless
« on: April 27, 2014, 11:03:39 PM »

I'd really like to know if this is in the pipeline or not, or could be considered for it. This would be a *very* valuable feature add for usage on remote machines. And its an issue that has come up again and is sadly making me reevaluate my entire workflow as I can not scale up to meet processing needs effectively.

General / Re: How about Canon EOS M
« on: August 08, 2013, 11:44:54 PM »
I've seen that actually, while it's cool and I really hope it matures right, it doesn't yet support RTK (software isn't done) and while I expect them to complete it, I still don't have a way to tie that in with the camera photos, or at least not at any better accuracy (due to using the granularity of the camera timestamp).

A follow up question, is does anyone know the protocol for the hotshoe, for feeding GPS data?

General / Re: How about Canon EOS M
« on: August 08, 2013, 12:01:26 PM »
What I don't know, is how well the GPS hotshoe works, which I'd be very intrested in knowing the performance of that. (Update rate to position and accuracy). We've been working with correlating photo location by establishing a time offset, and running through the plane log to find the correct location to match the timestamp. Been very intrested in the GPS, and was wondering if anyone else has tried it?

I can see wanting it, but I'd also like to be able to stick with the current blue squares. When checking for alignment/height issues possibly relating to geotagging it's very convenient to just see blue squares.

General / Re: How about Canon EOS M
« on: August 06, 2013, 09:26:29 AM »
I'm using the 22MM pancake. Do to the tiny size of the platform we are flying on I've never heard good things about IS actually. So we have't minded not having it, and we provide a 1-Axis Stabilizer mount for it, which really just serves to keep the camera nadir on that axis. But due to no IS and the huge reduction in weight made it worth it for us.

In magic lantern we have the delay set to 1 second on autofocus, which translates to an image roughly every 3 seconds aloft (due to relatively small change in the required focus) at smooth large jpegs with autofocus on single shot mode. If the focusing has gotten better then this time could go down. If you are willing to set focus manually first you can put the camera on continuous capture before starting magic lantern and get images at maximum capture rate, but it's impossible to get it to focus in this mode. Due to a bug in the autofocus algorthim + magic lantern, when it can't solve a focus problem (gives you a red square) it will stop capturing images, and to avoid this we use the delay timer, set it to 2 minutes which is adequate time to get the plane in the air and avoid any bad focus situation. (The camera sitting an inch away from the ground can't focus in our setup). As for endurance, we get at least an hour and a half of endurance without a problem. Due to normally flying 45 minute legs, I can't tell you where dead battery will be except that it is past an hour and a half. Obviously minimizing screen brightness is important on this, as well as making sure the camera wont poweroff before the intervelometer starts.

I didn't know about that on the firmware. That's good to know actually, as we are thinking about getting another one. We swapped from a higher resolution point and shoot to the EOS M and have been getting much better results due to the change in sensor size and the glass, both predictable results.

As far as external shuttering options go, I'm not aware of being able to connect it directly to the camera like that. It's possible I missed that, I never saw any mention of support for external shuttering, and before ordering the camera I had already settled on the ML solution and didn't investigate further (except checking out tethering which doesn't happen).

General / Re: TOP cam for low-altitude UAV
« on: July 29, 2013, 02:06:47 PM »
In the model you linked, was that created using a multi-rotor platform?

General / Re: How about Canon EOS M
« on: July 29, 2013, 02:06:03 PM »
Yeah, I keep hoping to see someone say it works, but haven't heard it yet.

The Canon Rebel SL1 might meet your needs. Pretty much the same specs as the EOS M, except it is a DLSR, same sensor (I think, its at least the same resolution and APS-C so it seems like a good bet). Its a bit large physical case, but as an actual DLSR I'd expect it to have tethering capabilities, although I haven't been able to test it yet. The SL1 is not quite at cheap as the EOS M has gotten recently but its around what I orginally paid to get an EOS M so it may be worth looking at.

General / Re: How about Canon EOS M
« on: July 27, 2013, 02:22:46 PM »
It did not add it for me, maybe it will show up in a future update. But at the moment you can have Magic Lantern in intervelometer taking many pictures, as soon as USB is connected all image capture stops immediately. GPhoto can't see anyway to capture photos, although it can pull images.

I really hope it gets it added at somepoint to Magic Lantern or officially supported.

General / Re: OS that you use
« on: July 26, 2013, 09:42:26 AM »
  • Fedora Core 19 - LXDE Spin
  • Debian Wheezy
  • Debian Sid

Tried windows early on, but the install of windows 7 was preventing exporting a orthophoto over 2GBs (would stop at exactly 2), took the same project file to Linux and it exported flawlessly. In general I'm primarily a Linux user and am more comfortable with that. (Especially if the ability to run without a GUI, and just pass in a python script was added, I brought it up once and never got any response past someone else wanting it  ::) )

Pages: [1] 2