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 - olihar

Pages: 1 ... 3 4 [5] 6 7 ... 9
61
General / Re: Drones
« on: May 11, 2012, 12:25:36 PM »
@Olihar:
Spam? That was my 16th post on this forum.
English isn?t my native language.
Don?t know whats your problem.
Make fun of me and be happy.

I think he referred to the link not being in English, but I could see how his post could be misunderstood.

@olihar: perhaps you should be more careful in regards to how you express yourself in the future? :P

I was not going against how you wrote the post, but I was stating the link you posted is in a different language then most people speak here on the board, but I guess its the same as all the other radio controlled mikrokopter forum boards who are marked as international tend to fade off to different languages.

@Kjellis85 I guess I should, but this link to me was a direct promote my own/my friends business kind of post, no real facts about hey I have 2 of these and I have managed to shoot these models from the air and it works great. But like you stated, you did get what I meant about the link being none relevant for a none German speaking person.

62
General / Re: Drones
« on: May 10, 2012, 07:53:46 PM »
If you are going to spam an english speaking forum board it would be the minimum requirement for the page to spam to be in english.

63
General / Photoscan Crash
« on: May 10, 2012, 02:42:09 AM »

I just got a crash with Photoscan, not idea the reason.

Code: [Select]
ARB_texture_non_power_of_two: Yes
ARB_vertex_buffer_object: Yes
OpenGLMaxTextureSize: 16384
OpenGLRenderer: GeForce GTX 580/PCIe/SSE2
OpenGLStereo: No
OpenGLVendor: NVIDIA Corporation
OpenGLVersion: 4.2.0
ProductName: PhotoScan Professional Edition
System: Windows 64bit
Throttleable: 1
URL:
Vendor: Agisoft
Version: 0.8.5.1423
XLog: OpenGL Vendor: NVIDIA Corporation
OpenGL Renderer: GeForce GTX 580/PCIe/SSE2
OpenGL Version: 4.2.0
Maximum Texture Size: 16384
Quad Buffered Stereo: not supported
ARB_vertex_buffer_object: supported
ARB_texture_non_power_of_two: supported
loaded project in 23.269 sec
Finished processing in 0.002 sec (exit code 1)
Finished processing in 1.409 sec (exit code 1)
tessellating mesh...done (1128474 -> 1128582 faces)
Mosaic size: 8242x11561
Estimating mosaic occupancy...
1 blocks used from 1x1 (100%)
Rendering mosaic...
rendering block
writing block
Finished processing in 6.207 sec (exit code 1)
Finished processing in 0 sec (exit code 1)
Finished processing in 1.149 sec (exit code 1)
tessellating mesh...done (1128474 -> 1128582 faces)
Mosaic size: 82425x115612
Estimating mosaic occupancy...
1 blocks used from 1x1 (100%)
Rendering mosaic...
rendering block
Error: aborted by user
Finished processing in 66.242 sec (exit code -1)
Finished processing in 0 sec (exit code 1)
Finished processing in 1.147 sec (exit code 1)
tessellating mesh...done (1128474 -> 1128582 faces)
Mosaic size: 82425x115612
Estimating mosaic occupancy...
1 blocks used from 1x1 (100%)
Rendering mosaic...
rendering block
writing block
Finished processing in 1454.17 sec (exit code 1)
Finished processing in 0 sec (exit code 1)
Finished processing in 1.154 sec (exit code 1)
tessellating mesh...done (1128474 -> 1128582 faces)
Mosaic size: 16485x23122
Estimating mosaic occupancy...
1 blocks used from 1x1 (100%)
Rendering mosaic...
rendering block
writing block
Finished processing in 14.751 sec (exit code 1)
initializing...
selected 8 cameras from 8 in 0.023 sec
Loading photos...
Reconstructing depth...
Using device: GeForce GTX 580, 16 compute units, 1535 MB global memory
  max work group size 1024
  max work item sizes [1024, 1024, 64]
  max mem alloc size 383 MB
  max workgroup size c1: 1024 c3: 1024 zero: 1024 hamming: 1024 filter: 1024 box: 1024
  max workgroup size zero: 1024 costs: 1024 b1: 1024 bn: 1024 wta: 1024 transpose: 1024
estimating 1280x8655x128 disparity using 256x255x128 tiles, offset 0
estimating 3488x8658x128 disparity using 250x255x128 tiles, offset 0
timings: rectify: 0.444 disparity: 10.119 borders: 0.229 filter: 0.225 fill: 0
estimating 3488x8672x96 disparity using 250x256x96 tiles, offset 0
timings: rectify: 0.978 disparity: 19.398 borders: 0.436 filter: 0.79 fill: 0
estimating 1248x8654x160 disparity using 250x255x160 tiles, offset 0
timings: rectify: 0.326 disparity: 8.436 borders: 0.122 filter: 0.299 fill: 0
estimating 1248x8656x160 disparity using 250x255x160 tiles, offset 0
timings: rectify: 0.472 disparity: 24.108 borders: 0.391 filter: 0.645 fill: 0
estimating 2896x5618x128 disparity using 264x256x128 tiles, offset 0
timings: rectify: 0.291 disparity: 8.909 borders: 0.147 filter: 0.298 fill: 0
estimating 3488x8657x128 disparity using 250x255x128 tiles, offset 0
timings: rectify: 0.233 disparity: 15.296 borders: 0.281 filter: 0.342 fill: 0
estimating 3952x8353x128 disparity using 264x254x128 tiles, offset 0
timings: rectify: 0.756 disparity: 19.425 borders: 0.378 filter: 0.674 fill: 0
estimating 3456x8657x128 disparity using 247x255x128 tiles, offset 0
timings: rectify: 0.878 disparity: 19.462 borders: 0.541 filter: 0.667 fill: 0
estimating 2896x5614x128 disparity using 264x256x128 tiles, offset 0
timings: rectify: 0.434 disparity: 31.469 borders: 0.259 filter: 0.593 fill: 0
estimating 3536x8657x96 disparity using 253x255x96 tiles, offset 0
timings: rectify: 0.468 disparity: 10.088 borders: 0.194 filter: 0.437 fill: 0
estimating 3920x8256x128 disparity using 262x258x128 tiles, offset 0
timings: rectify: 0.471 disparity: 24.387 borders: 0.417 filter: 0.605 fill: 0
estimating 3472x8664x96 disparity using 248x255x96 tiles, offset 0
timings: rectify: 0.636 disparity: 20.342 borders: 0.255 filter: 0.64 fill: 0
estimating 2928x5655x128 disparity using 267x258x128 tiles, offset 0
timings: rectify: 0.388 disparity: 10.49 borders: 0.228 filter: 0.416 fill: 0
timings: rectify: 0.467 disparity: 22.251 borders: 0.755 filter: 0.736 fill: 0
estimating 3536x8656x96 disparity using 253x255x96 tiles, offset 0
estimating 1280x8655x128 disparity using 256x255x128 tiles, offset 0
timings: rectify: 0.242 disparity: 9.911 borders: 0.2 filter: 0.224 fill: 0
estimating 1248x8655x192 disparity using 250x255x192 tiles, offset 0
timings: rectify: 0.917 disparity: 17.577 borders: 0.375 filter: 0.865 fill: 0
estimating 1280x8677x96 disparity using 256x256x96 tiles, offset 0
timings: rectify: 0.293 disparity: 6.356 borders: 0.14 filter: 0.261 fill: 0
estimating 3488x8660x128 disparity using 250x255x128 tiles, offset 0
timings: rectify: 0.166 disparity: 18.199 borders: 0.185 filter: 0.209 fill: 0
estimating 3456x8656x128 disparity using 247x255x128 tiles, offset 0
timings: rectify: 0.937 disparity: 19.576 borders: 0.361 filter: 0.738 fill: 0
estimating 3488x8657x128 disparity using 250x255x128 tiles, offset 0
timings: rectify: 0.461 disparity: 28.322 borders: 0.337 filter: 0.528 fill: 0
estimating 2912x5653x128 disparity using 265x257x128 tiles, offset 0
timings: rectify: 0.903 disparity: 19.52 borders: 0.556 filter: 0.809 fill: 0
estimating 1248x8656x192 disparity using 250x255x192 tiles, offset 0
timings: rectify: 0.199 disparity: 15.841 borders: 0.287 filter: 0.312 fill: 0
estimating 3936x8351x128 disparity using 263x254x128 tiles, offset 0
timings: rectify: 0.314 disparity: 9.872 borders: 0.189 filter: 0.306 fill: 0
estimating 1280x8677x96 disparity using 256x256x96 tiles, offset 0
timings: rectify: 0.3 disparity: 6.282 borders: 0.137 filter: 0.307 fill: 0
estimating 3472x8663x96 disparity using 248x255x96 tiles, offset 0
timings: rectify: 0.426 disparity: 31.791 borders: 0.256 filter: 0.588 fill: 0
timings: rectify: 0.873 disparity: 17.292 borders: 0.509 filter: 0.636 fill: 0
estimating 3488x8673x96 disparity using 250x256x96 tiles, offset 0
timings: rectify: 0.732 disparity: 16.006 borders: 0.396 filter: 0.716 fill: 0
finished depth reconstruction in 279.283 seconds
Device 1 performance: 198.8 million samples/sec (CPU)
Device 2 performance: 293.863 million samples/sec (GeForce GTX 580)
Total performance: 492.663 million samples/sec
Generating mesh...
Grid size: 2321 x 2321
Tree depth: 12
48681666 points extracted
samples: 48681666
setting sample weights
adding points
extracting mesh
33433029 faces extracted
filtering mesh (33433029 -> 33433029)
Finished processing in 634.538 sec (exit code 1)
Finished processing in 0 sec (exit code 1)
initializing...
selected 8 cameras from 8 in 0.072 sec
Loading photos...
Reconstructing depth...
Using device: GeForce GTX 580, 16 compute units, 1535 MB global memory
  max work group size 1024
  max work item sizes [1024, 1024, 64]
  max mem alloc size 383 MB
  max workgroup size c1: 1024 c3: 1024 zero: 1024 hamming: 1024 filter: 1024 box: 1024
  max workgroup size zero: 1024 costs: 1024 b1: 1024 bn: 1024 wta: 1024 transpose: 1024
estimating 2496x17310x160 disparity using 250x255x160 tiles, offset 0
estimating 6928x17316x160 disparity using 257x255x160 tiles, offset 0
timings: rectify: 1.48 disparity: 48.366 borders: 0.518 filter: 0.936 fill: 0
estimating 6960x17345x128 disparity using 258x256x128 tiles, offset 0
timings: rectify: 3.636 disparity: 95.573 borders: 1.85 filter: 3.698 fill: 0
estimating 2480x17309x256 disparity using 248x255x256 tiles, offset 0
timings: rectify: 1.353 disparity: 51.166 borders: 0.893 filter: 1.311 fill: 0
estimating 6928x17321x160 disparity using 257x255x160 tiles, offset 0
timings: rectify: 2.2 disparity: 118.013 borders: 2.029 filter: 3.392 fill: 0
estimating 2480x17312x256 disparity using 248x255x256 tiles, offset 0
timings: rectify: 0.777 disparity: 69.229 borders: 0.52 filter: 0.848 fill: 0
estimating 6928x17314x160 disparity using 257x255x160 tiles, offset 0
timings: rectify: 3.012 disparity: 97.523 borders: 1.935 filter: 3.55 fill: 0
estimating 6928x17314x160 disparity using 257x255x160 tiles, offset 0
timings: rectify: 4.275 disparity: 95.944 borders: 2.309 filter: 3.206 fill: 0
estimating 2464x17313x320 disparity using 247x255x320 tiles, offset 0
timings: rectify: 2.258 disparity: 142.542 borders: 1.302 filter: 2.303 fill: 0
estimating 6864x17314x160 disparity using 255x255x160 tiles, offset 0
timings: rectify: 1.217 disparity: 68.954 borders: 0.778 filter: 1.163 fill: 0
estimating 5776x11237x192 disparity using 252x256x192 tiles, offset 0
timings: rectify: 1.469 disparity: 60.768 borders: 1.116 filter: 2.134 fill: 0
estimating 7824x16706x160 disparity using 253x258x160 tiles, offset 0
timings: rectify: 2.16 disparity: 142.643 borders: 1.3 filter: 2.234 fill: 0
estimating 5776x11229x192 disparity using 252x256x192 tiles, offset 0
timings: rectify: 0.785 disparity: 82.568 borders: 0.672 filter: 1.294 fill: 0
timings: rectify: 3.034 disparity: 104.565 borders: 1.231 filter: 2.314 fill: 0
estimating 6992x17314x96 disparity using 259x255x96 tiles, offset 0
estimating 7776x16513x160 disparity using 260x255x160 tiles, offset 0
timings: rectify: 3.421 disparity: 70.538 borders: 2.236 filter: 3.253 fill: 0
estimating 6912x17328x128 disparity using 256x255x128 tiles, offset 0
timings: rectify: 1.74 disparity: 145.929 borders: 0.854 filter: 1.952 fill: 0
estimating 5824x11310x192 disparity using 254x258x192 tiles, offset 0
timings: rectify: 3.572 disparity: 78.644 borders: 2.218 filter: 3.579 fill: 0
estimating 2496x17311x160 disparity using 250x255x160 tiles, offset 0
timings: rectify: 1.399 disparity: 34.85 borders: 0.894 filter: 1.385 fill: 0
estimating 2464x17310x320 disparity using 247x255x320 tiles, offset 0
timings: rectify: 0.798 disparity: 103.144 borders: 0.776 filter: 1.314 fill: 0
estimating 6992x17312x96 disparity using 259x255x96 tiles, offset 0
timings: rectify: 1.184 disparity: 69.271 borders: 0.817 filter: 1.249 fill: 0
estimating 6848x17312x160 disparity using 254x255x160 tiles, offset 0
timings: rectify: 2.28 disparity: 90.262 borders: 1.33 filter: 2.233 fill: 0
estimating 2528x17354x128 disparity using 253x256x128 tiles, offset 0
timings: rectify: 0.697 disparity: 39.758 borders: 0.498 filter: 0.914 fill: 0
timings: rectify: 3.624 disparity: 93.497 borders: 1.98 filter: 3.113 fill: 0
estimating 2528x17355x128 disparity using 253x256x128 tiles, offset 0
estimating 5808x11306x192 disparity using 253x257x192 tiles, offset 0
timings: rectify: 1.51 disparity: 41.4 borders: 0.541 filter: 0.852 fill: 0
estimating 6912x17327x128 disparity using 256x255x128 tiles, offset 0
timings: rectify: 1.279 disparity: 60.363 borders: 1.302 filter: 2.059 fill: 0
estimating 7824x16702x160 disparity using 253x257x160 tiles, offset 0
timings: rectify: 2.177 disparity: 112.392 borders: 1.328 filter: 2.284 fill: 0
estimating 6960x17346x128 disparity using 258x256x128 tiles, offset 0
timings: rectify: 3.241 disparity: 103.193 borders: 1.029 filter: 2.524 fill: 0
timings: rectify: 2.213 disparity: 116.088 borders: 1.228 filter: 2.649 fill: 0
finished depth reconstruction in 1379.28 seconds
Device 1 performance: 211.268 million samples/sec (CPU)
Device 2 performance: 307.291 million samples/sec (GeForce GTX 580)
Total performance: 518.559 million samples/sec
Generating mesh...
Grid size: 4643 x 4643
Tree depth: 13
188291105 points extracted
samples: 188291105
setting sample weights
adding points
extracting mesh
131041185 faces extracted
filtering mesh

This report also contains technical information about the state of the application when it crashed.



64
General / Creation of DEM
« on: May 09, 2012, 08:07:30 PM »
Ok so I am trying to understand the DEM export better.

When I export DEM does it recompute the model, does it not use the 3D model already reconstructed.

Reason I am asking I am doing some tests and it takes around 45GB of RAM to export a fairly small area, it is like it is reprocessing the model.

Edit: So if I put lower value then default given value it will do this heavy process.

65
Feature Requests / Re: bounding box center by clicking
« on: May 09, 2012, 05:23:45 PM »
Noone think this might work?

66
General / omega phi kappa to Yaw Pitch Roll
« on: May 07, 2012, 06:30:06 PM »
Hey gang

If I have coordinates for each image along with omega phi kappa How would I go by into converting them into Yaw Pitch Roll like Photoscan supports?

And what order is it in.

Omega > Roll
Phi > Pitch
Kappa > Yaw

Or do I have it wrong?

Can I simply insert the omega phi kappa values into Photoscan maybe without anything else.

67
General / Re: VBO support
« on: May 07, 2012, 12:47:39 AM »
Well turning this one makes Photoscan a lot faster browsing around the model, without it can barely move, with it, it does not even have to think.

68
General / VBO support
« on: May 06, 2012, 05:46:58 AM »
So what is this VBO support, there is no info when moused over, nor is there anything in the manuals about it.

Google says Vertex Buffer Object in relations to OpenGL.

Does it force the graphics card to keep more in memory, or what is it all about?

69
General / Re: ply export
« on: May 06, 2012, 05:45:14 AM »
ok, so I just tried this... why does it have to reconstruct the model again? As I just finished a model on high inside the program, then I wanted to have a look at the point cloud in high setting, but exporting point cloud on high has to rebuild the whole model again... Would it not be possible to use the already processed data.

71
General / ply export
« on: April 30, 2012, 04:03:59 AM »
I am wondering is it possible to export only the points with color in for example ply or just raw txt format from Photoscan, I can only export to ply including faces. As well the file is in binary, so I have to open it in meshlab for example and save as ASCII, edit in text editor and then I have a clean vertex only file. However problem is what if you have million or even billions of points, Meshlab has a hard time opening it.

Therefore I would love an output that only outputs like so. for example a ply file. x, y, z, normals and color.

property float x
property float y
property float z
property float nx
property float ny
property float nz
property uchar red
property uchar green
property uchar blue
property uchar alpha


alpha is of course extra and does not have to be there.

I cant imagine this being hard to implement, or maybe I am missing something and it is already there.

I am thinking this because it can often be easier to visualize billions of points then 100 million faces.

P.s. I am actually wondering about the reconstruction process as well, when the GPU step has finished does that not indeed return a simple point cloud data? could then the last steps be skipped if one would only be after the point cloud data, or... that data could be processed a little later, to brake down the reconstruction steps.

72
Feature Requests / bounding box center by clicking
« on: April 30, 2012, 02:56:15 AM »
I have been wondering as I think moving the bounding box is a little clunky... could there be a way to make this a little easier... for example you could click on the model or point and make that the center of the bounding box. As well a numeric value for the bounding box could help a lot with such a click function where you assign how far on each axis it will go from the click you just did.

The box could as well be used to change rotation and orientation of the bounding box, say you want to flip the bounding box 180° or make it more accurate for height field reconstruction.

I hope I make myself clear on this. And I do think this would be a very helpful function.

Little extra, the numeric window could also have a function where you tell it to move x far from its current position, for example if you want to reconstruct a planar model in passes, you can simply move the box numerically instead of manually.

73
General / Re: Cache directory in Photoscan
« on: April 11, 2012, 12:33:08 AM »
Ok so it finished using Height Field... However... it just processed about 1/4 of the data and left the rest empty inside the bounding box.

I saw that it had extracted just under 6 Billion points in the reconstruction step, is there some limitation on how much detail one can get. As well I set the mesh to 50 million, is there a limit there as well or did I not set it high enough to be able to cover the whole area?

74
General / Re: Cache directory in Photoscan
« on: April 10, 2012, 04:49:33 PM »
Hello olihar,

And are you using Height Field object type? Memory consumption should be less than in Arbitrary mode.

Testing Height field now... Seems to use a lot as well... It went through the reconstruction just fine as before, I will await the Generation of the Mesh, as again it seems to be just a single thread process it will be slow.

Its using 110GB right now in the beginning of the process.



Is there any way this step could be multi-threaded in the future?

P.s. can it be that Photoscan is not letting go of the RAM in the reconstruction process when it goes into the Generating mesh step, or does it have to keep everything in RAM?

75
General / Re: Cache directory in Photoscan
« on: April 09, 2012, 08:23:17 PM »
Hello olihar,

What type of data are you processing and what parameters are you using?

Its data from UltraCamX Prime...
196 megapixels (17,310 across track x 11,310 along track),
 PDF here for info.
http://ecogis.pl/PDF/UCXP.pdf

I can run the data just fine on Medium using 10 million faces. But High and Ultra do not seem to be possible, as I get the error above. Its like it manages to cache to a certain point then it gets the error.

The bottom of the logs using medium, just so its been posted here.

Code: [Select]
Linear system solved in 6297.36s
Got Iso-value in 836.567s
Iso-value -18255.5
Normal Size: 1655.4 MB
Set 82933188 root positions
Got triangles in 47892.9s
filtering mesh (165866237 -> 165864653)
decimating mesh (165864653 -> 10000000)
Decimating mesh...
heap cleanup: 273676713 -> 231045734
heap cleanup: 273676715 -> 204805798
heap cleanup: 273676717 -> 153848387
Performance: 404.585 million samples/sec
Finished processing in 60851.1 sec (exit code 1)
saved project in 21.638 sec
>>>

Pages: 1 ... 3 4 [5] 6 7 ... 9