1
Python and Java API / Re: Export CRI point cloud
« on: July 28, 2017, 11:25:08 AM »
I have exactly the same question!
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.
Pages: [1] 2
2
Bug Reports / Gradual Selection in 1.2.3
« on: March 04, 2016, 09:47:57 AM »
Hello,
I'm currently working on a project with 2.500 images. The alignment of all images (geotagged) works fine, it lead to 17.000.000 Tie points in the sparse point cloud.
Now I would like to keep only those points that match at least in three images.
Gradual selection is the solution since years.
But in Version 1.2.3 my system crashes.
Win 7, 128 GB RAM (13 GB in use)
Two GTX 780ti
Dual Xeon 2620
I think, the machine has enough power to select the points, but agisoft hangs in a "loop", reacts only very slowly on any command and I cannot finish the gradual selection tool. (deleting button doesn't work, photoscan doesn't react).
Limiting the number of tie point (4.3 Mio) works fine on the same machine with identical data.
Is there a limit for the gradual selection tool?
Best regards
Ansgar
I'm currently working on a project with 2.500 images. The alignment of all images (geotagged) works fine, it lead to 17.000.000 Tie points in the sparse point cloud.
Now I would like to keep only those points that match at least in three images.
Gradual selection is the solution since years.
But in Version 1.2.3 my system crashes.
Win 7, 128 GB RAM (13 GB in use)
Two GTX 780ti
Dual Xeon 2620
I think, the machine has enough power to select the points, but agisoft hangs in a "loop", reacts only very slowly on any command and I cannot finish the gradual selection tool. (deleting button doesn't work, photoscan doesn't react).
Limiting the number of tie point (4.3 Mio) works fine on the same machine with identical data.
Is there a limit for the gradual selection tool?
Best regards
Ansgar
3
Bug Reports / Re: Calculating vertex colors
« on: February 22, 2016, 06:41:39 PM »
The link to the tar archive is:
http://download.agisoft.com/photoscan-pro_1_2_3_amd64.tar.gz
just replace the version number with your last stable release.
Since I have also Problems with 1.2.0, which is the last stable release?
@ Alexey: How do I have to understand your comment? Do all Linux Users have to wait no for month? Is that your new politics? Should we move to windooooze?
http://download.agisoft.com/photoscan-pro_1_2_3_amd64.tar.gz
just replace the version number with your last stable release.
Since I have also Problems with 1.2.0, which is the last stable release?
@ Alexey: How do I have to understand your comment? Do all Linux Users have to wait no for month? Is that your new politics? Should we move to windooooze?
4
General / Re: Calculating vertex colors hangs
« on: February 06, 2016, 07:21:06 PM »
I have the same problem... ubuntu 14.04 with NVIDIA
5
General / Re: Tetracam adc multispectral camera processing AGAIN
« on: September 09, 2015, 03:11:05 PM »
Hello,
I used that camera and even got bad results. I think it's due to the global shutter of that camera. (Cheap cameras like GoPro have no mechnaical shutter). And also that ADC Camera has no mechanical shutter.
Sit down on a revolving chair (perhaps you already sit on one).
Take the ADC and prepare to take some shoots.
rotate the chair while sitting on it
take some pictures
stop after a few rotations
take a look to the photos.
Things (like your monitor) with perpendicular lines will not have vertival line in the image.
Everything will be distorted between the vertical lines of the image, because the sensor is read line by line (this means: rolling shutter).
Perhaps you can send us such a photo to the forum as a nice example for "cameras the world doesn't need".
I put away my ADC to the garbage a few week ago. My vendor in Germany say "it's a feature".
And - NO - theres no chance to compensate the effect, your UAV will always move in some kind of direction, it's not a matter of microvibrations (take a revolving chair, make the test....)
I used that camera and even got bad results. I think it's due to the global shutter of that camera. (Cheap cameras like GoPro have no mechnaical shutter). And also that ADC Camera has no mechanical shutter.
Sit down on a revolving chair (perhaps you already sit on one).
Take the ADC and prepare to take some shoots.
rotate the chair while sitting on it
take some pictures
stop after a few rotations
take a look to the photos.
Things (like your monitor) with perpendicular lines will not have vertival line in the image.
Everything will be distorted between the vertical lines of the image, because the sensor is read line by line (this means: rolling shutter).
Perhaps you can send us such a photo to the forum as a nice example for "cameras the world doesn't need".
I put away my ADC to the garbage a few week ago. My vendor in Germany say "it's a feature".
And - NO - theres no chance to compensate the effect, your UAV will always move in some kind of direction, it's not a matter of microvibrations (take a revolving chair, make the test....)
6
General / Re: Cluster Processing - RAM consumption of nodes
« on: September 01, 2015, 03:06:30 PM »
Hello Alexey,
It is the "filtering depth Images" - step in the command "Build dense Cloud".
The first step is done by GPU and works pretty well on all nodes.
Afterwards, on the CPU, the filtering is done.
And during this process, right at the beginning, each node loads it's data into RAM (I think the depth Images) and starts swapping due to the high memory consumption (54GB, 32 available).
It is the "filtering depth Images" - step in the command "Build dense Cloud".
The first step is done by GPU and works pretty well on all nodes.
Afterwards, on the CPU, the filtering is done.
And during this process, right at the beginning, each node loads it's data into RAM (I think the depth Images) and starts swapping due to the high memory consumption (54GB, 32 available).
7
General / Re: Cluster Processing - RAM consumption of nodes
« on: September 01, 2015, 08:00:13 AM »
The project has 1000 oblique images, each 15 MPix.
Split in chunks is no option due to the oblique imagery.
As I mentioned, i tried to process it as cluster.
In fine level task distribution, the server creates 60 tasks.
However, each task needs about 54 GB to process the data.
My cluster nodes have only 32 GB per node, so they start swapping.
I also tried to switch off the swap file. The result is the same, but now the nodes crashed instead of swapping (which is expected).
So my question is: Does the server take the node's memory into account, when he is defining the tasks? Can I increase or decrease the number of tasks? Or can I specify the "max memory" for a task in cluster processing?
It's no "windows" issue, a linux setup of my cluster showed the same behaviour...
Split in chunks is no option due to the oblique imagery.
As I mentioned, i tried to process it as cluster.
In fine level task distribution, the server creates 60 tasks.
However, each task needs about 54 GB to process the data.
My cluster nodes have only 32 GB per node, so they start swapping.
I also tried to switch off the swap file. The result is the same, but now the nodes crashed instead of swapping (which is expected).
So my question is: Does the server take the node's memory into account, when he is defining the tasks? Can I increase or decrease the number of tasks? Or can I specify the "max memory" for a task in cluster processing?
It's no "windows" issue, a linux setup of my cluster showed the same behaviour...
8
General / Cluster Processing - RAM consumption of nodes
« on: August 31, 2015, 05:59:08 PM »
Hello,
I have got a project, that ends up in 60 packages (nodes) and the processing on each single node consumes 54 GB RAM. So the node start swapping (windows or linux with the same behaviour).
Is ist possible to define the max amount of ram, that is allowed for a node in a cluster?
Fine level task distribution is already activated for all steps...
Best regards
Ansgar
I have got a project, that ends up in 60 packages (nodes) and the processing on each single node consumes 54 GB RAM. So the node start swapping (windows or linux with the same behaviour).
Is ist possible to define the max amount of ram, that is allowed for a node in a cluster?
Fine level task distribution is already activated for all steps...
Best regards
Ansgar
9
General / NVIDIA Driver Version for Linux - SOLVED
« on: August 19, 2015, 05:47:25 PM »
Hello,
I tried to install Agisoft on a new machine.
Version 1.1.6 works well with NVIDIA GTX 980ti and Windows Driver 353.30 on Windows 8.1 Pro.
Now I would like to switch to Linux (for several reasons like clustering etc.)
Working with 12.04.5 LTS and Driver Version 331 works well on another machine with a GTX 780ti
Installing 14.04 LTS with NVIDIA 340 (to support 980ti) doesn't work.
Agisoft doesn't show my cards in the "OpenCL" tab in preferences.
Switshing to Agisoft 1.2 pre release does't help
---------------------------------------------------------------------------
Just use this link, now it works well, with 750ti, 780ti and 980ti
http://askubuntu.com/questions/464354/update-nvidia-drivers-with-xorg-edgers-ppa
---------------------------------------------------------------------------
I tried to install Agisoft on a new machine.
Version 1.1.6 works well with NVIDIA GTX 980ti and Windows Driver 353.30 on Windows 8.1 Pro.
Now I would like to switch to Linux (for several reasons like clustering etc.)
Working with 12.04.5 LTS and Driver Version 331 works well on another machine with a GTX 780ti
Installing 14.04 LTS with NVIDIA 340 (to support 980ti) doesn't work.
Agisoft doesn't show my cards in the "OpenCL" tab in preferences.
Switshing to Agisoft 1.2 pre release does't help
---------------------------------------------------------------------------
Just use this link, now it works well, with 750ti, 780ti and 980ti
http://askubuntu.com/questions/464354/update-nvidia-drivers-with-xorg-edgers-ppa
---------------------------------------------------------------------------
10
General / Re: 980GTX Ti
« on: July 15, 2015, 10:04:26 PM »
Hi,
the 980 ti work very well under win 8.1 pro with nvidia driver 353.30
33% better than my good old GTX Titan
Cheers
the 980 ti work very well under win 8.1 pro with nvidia driver 353.30
33% better than my good old GTX Titan
Cheers
11
General / Re: Processed my aerial model, How Can I do a fly through video of it?
« on: February 01, 2015, 09:10:50 PM »
blender is also a nice rendering software also available linux.
12
General / Re: GCP error low - high errors on DEM
« on: February 01, 2015, 09:07:57 PM »Hello Alexey,
It has been brought to my attention that the lens that I used is most likely the culprit of the deformation, thanks to the lens image stabilization that was turned ON.
In the end, GIS works by "Garbage In, Garbage Out". I suspect new photographs will have to be taken.
Yes, this is exactly the point. Your principal point in camera calibration process could not be estimated correctly if you have turned on an image stabilizer. Even the autofocus should be turned off, but has not that terrible effect like the image stabilizer.
You'll have to take the images again.
13
General / Re: Issue getting correct orhtomap
« on: February 01, 2015, 08:58:54 PM »
I do not think that the problem is only the absence of gcp.
First of all, the dense point cloud show a lot of noise.
Too much for my humble opinion.
The surface is very light and has not much texture. This makes it difficult to match identical points in different images.
And this process depends highly on you image wuality. Which camera are you using? Perhaps this could be an issue?
First of all, the dense point cloud show a lot of noise.
Too much for my humble opinion.
The surface is very light and has not much texture. This makes it difficult to match identical points in different images.
And this process depends highly on you image wuality. Which camera are you using? Perhaps this could be an issue?
14
General / Re: Best camera for Photoscan?
« on: January 20, 2015, 08:23:53 PM »Looks like my main problem with Photoscan is not enough light for f6- f8 usually and at least 1\200 shuttler speed for having crisp images under cloudy/ambient light i need for good shadeless textures .
So I wonder if I really need a big camera and optic. What I get of general point-and-shot cameras is somehow too noisy geometry.
Since 5 years I've been using Sigma compact cameras with a foveon chip.
You will never find any better MTF for about 500 Euros.
I tested a Phase One iQ180 and a Sigma DP1 Merrill.
Copare the attached meshes...
Of course there are some disadvantages working with the Merrill: To store the RAW data, it takes 15 seconds and the noise is terrible at ISO ratings bigger than 200.
However, I would always take the Merrill for UAV-missions, the Phase One for terrestrial Photogrammetry.
15
General / Re: Tetracam multispectral imagery processing
« on: July 02, 2014, 12:13:38 PM »
Dear Martin,
it is as it is, the tetracam has a rolling shutter. Any movement of a pamera platform will distort the image.
If your camera platform is moving fast (plane?), you will get a somewhat "stretched" image in one direction (e.g. y - axis of image coordinate system) and a very small distortion in the other direction (e.b. x axis). And PhotoScan will handle that different image scale in x and y within the camera calibration. That why some datasets look "good".
They also look "acceptable" in some cases due to the low resolution of that camera.
But just take the ADC in your hands, sit down on an office chair, rotate yourself slowly and hold the adc 50 cm away from your breast (z-axis). You'll never get undistorted images.
It's not an issue of "low cost uav", it's an issue of CMOS...
Best regards
Ansgar
it is as it is, the tetracam has a rolling shutter. Any movement of a pamera platform will distort the image.
If your camera platform is moving fast (plane?), you will get a somewhat "stretched" image in one direction (e.g. y - axis of image coordinate system) and a very small distortion in the other direction (e.b. x axis). And PhotoScan will handle that different image scale in x and y within the camera calibration. That why some datasets look "good".
They also look "acceptable" in some cases due to the low resolution of that camera.
But just take the ADC in your hands, sit down on an office chair, rotate yourself slowly and hold the adc 50 cm away from your breast (z-axis). You'll never get undistorted images.
It's not an issue of "low cost uav", it's an issue of CMOS...
Best regards
Ansgar
Hi Community,
today was helping a member of this community with his ADC data. I had no Problems to create a perfect Ortho Mosaik with Agisoft PhotoScan.
Attached some Screen shots showing the steps ...
@AngarG: No rolling shutter :-)
Again: If some others of you need my help with ADC data, please get in touch with me !
Cheers
Martin
Pages: [1] 2
