Topics

31

Feature Requests / Batch/Script processing quietly in background

« on: October 21, 2014, 11:20:06 PM »

I wish PhotoScan wouldn't keep stealing focus from the app i am working in when I'm running a batch or python scripted process. Right now I am running a script to export around 100 orthophotos, each of which takes only 5 seconds, but means I've had to click in this textbox more than 10 times so far just to continue writing this short post!

If there was an option to run in background like winrar or something like that it would extend my life expectancy i'm sure. Clicking minimize currently only works until the next progress dialog pops up.

32

Feature Requests / Retrospective Depth Filtering

« on: July 10, 2014, 09:41:05 PM »

Further to this request http://www.agisoft.ru/forum/index.php?topic=1221.0, would it be possible to have an option to perform depth filtering after the dense cloud has been calculated and displayed?

Kind of like the existing gradual selection filters for the sparse cloud.

I just tried building my dense cloud with depth filtering set to mild for the first time(!) and was surprised how many extra valid points I seemed to get compared to aggressive filtering.

If it had all been garbage however it would have been nice to move a slider to remove the points that would have been filtered, without having to run the dense cloud calculation over again.

33

General / Tilt-shift lens

« on: April 09, 2014, 03:20:59 PM »

Hi there,

Does anyone have an idea about whether a tilt shift lens would work with photoscan?

I was just thinking it would be preferable to use one when scanning tall buildings from ground level, to maintain resolution at the higher levels, but never actually used, touched or seen one, so could be completely misguided!

Thanks!

James

34

Feature Requests / colour correction options

« on: January 16, 2014, 07:20:58 PM »

sometimes the colour correction gets it wrong and i get a big splodge of red or green etc on my model, and it can be hard to tell where it came from.

1. assuming colour correction is applied to the whole photo, it would be nice to toggle it on the thumbnails in the photos pane to see what went wrong.

2. also sometimes it seems to have been thrown off by part of the image that was masked out, so maybe an option to only colour correct based on non masked areas?

3. finally an option to save the colour correction values so they can be reused for subsequent texture generation, if that is a possibility?

in general the colour correction is great and makes a huge difference compared to before it was available.

james

35

Feature Requests / Align model to bounding box/Align box to model.

« on: October 18, 2013, 06:50:31 PM »

Sometimes it seems that the arbitrary initial coordinate system is rather random, whereas the bounding box is well aligned to the model. Sometimes once the model has been georeferenced the bounding box might be at a funny orientation. It would be great to be able to snap the coordinate system to the bounding box orientation and vice versa.

36

Feature Requests / Point cloud display size

« on: October 18, 2013, 06:48:09 PM »

I wish, I wish I could make the points display as single pixels instead of chunky blocks, particularly in dense cloud mode!

37

Feature Requests / Import mask - additive (or subtractive?)

« on: October 18, 2013, 06:46:53 PM »

It would be great when importing mask (either from file or from model) to have an option to add to or subtract from existing masks, rather than always overwrite.

38

General / Scale Bars vs GCPs

« on: October 18, 2013, 12:07:33 PM »

I recently scanned an interior space using photoscan's coded targets (which are amazing and all detected perfectly, but beware of reflections in windows and mirrors!) and a tape and handheld laser distometer to measure between them to generate a 'network' of 'triangulated' scale bars. After optimising the errors were very low, comparable to when using a total station to get full xyz coords for each target.

Assuming my scale bars were accurate to within a few mm, is there any reason why this approach would be less robust than acquiring full xyz coords for each marker? I assumed that by obtaining measurements between 'triangles' of markers, the network would be constrained to optimise to a fairly good solution.

I understand that this approach will not give me a good 'up' direction, but if i'd thought of it i could have helped this by using three markers on the floor.

Any thoughts? Thanks!

James

39

Feature Requests / Meshlab MLP Export for Cameras

« on: October 04, 2013, 07:17:27 PM »

Now meshlab has fun raster layer functions, it would be good to be able to export cameras as an MLP file.

This kind of format seems to work, although I don't claim to understand all the numbers!

Code: [Select]

<!DOCTYPE MeshLabDocument>
<MeshLabProject>
 <RasterGroup>
  <MLRaster label="DSC_5227.JPG">
   <VCGCamera TranslationVector="-3.12203 42.8511 -41.8889 1" LensDistortion="0 0" ViewportPx="4928 3264" PixelSizeMm="1 1" CenterPx="2464 1632" FocalMm="11207" RotationMatrix="0.997452 0.0712106 -0.00441334 0 0.00968197 -0.073812 0.997225 0 0.0706873 -0.994727 -0.0743134 0 0 0 0 1 "/>
   <Plane semantic="" fileName="DSC_5227.JPG"/>
  </MLRaster>
  <MLRaster label="DSC_5228.JPG">
   <VCGCamera TranslationVector="-2.33531 43.1249 -41.881 1" LensDistortion="0 0" ViewportPx="4928 3264" PixelSizeMm="1 1" CenterPx="2464 1632" FocalMm="11207" RotationMatrix="0.998128 0.0611126 -0.00248223 0 0.00626142 -0.0617261 0.998073 0 0.0608417 -0.99622 -0.0619932 0 0 0 0 1 "/>
   <Plane semantic="" fileName="DSC_5228.JPG"/>
  </MLRaster>
etc
 </RasterGroup>
</MeshLabProject>
This was created by exporting cameras from PS as bundler .out and manually creating the camera list file to import to meshlab, so it is possible to do it currently, and not important, but might be a nice feature.

40

Python and Java API / valid/invalid matches between image pairs

« on: July 24, 2013, 05:56:43 PM »

am i right in thinking that matching points between image pairs are not currently accessible though the python api?

i was hoping to be able to do some filtering based on ratio of valid to invalid matches or similar on a particularly messy dataset i have to work on.

james

41

Bug Reports / Single camera UV mapping and occlusion

« on: June 11, 2013, 07:42:32 PM »

It seems that when you select single camera mapping mode, it is sometimes hit and miss whether faces are UV mapped if they are partially occluded from the selected camera position.

Image below shows 2 partially occluded faces (left) are UV mapped (right) but the face between them has not been.

42

Feature Requests / Filter photos by enabled/disabled status

« on: June 10, 2013, 04:45:24 PM »

I'd love to be able to hide photos once I have disabled them. I seem to be disabling ~90% of my photos before building texture most of the time.

Possibly even better would be to dim/grey out disabled photos in the photos pane but leave visible.

It is hard to tell at a glance which photos are disabled just by the little red disabled icon, especially when markers are selected and you also have the marker icon on each photo, disabled or otherwise.

43

Feature Requests / Visual display of 'per image' marker error

« on: June 10, 2013, 03:48:58 PM »

When you position a marker it turns green and photoscan's inferred position (grey flag) disappears. I would like an option to leave it visible or some other way of showing the disparity i.e. error bars, numeric x,y pixel error etc.

If I am right in thinking (in my ever simplistic way) that the error values shown in ground control are based on the difference between where the markers have been placed and where photoscan thinks they should be, then it would be really useful to know where it thinks they should be.

Sometimes I stare at a project far too long trying to work out where an error is originating from and eventually realise it is because I put a marker in a stupid place. Most of the time photoscan's estimated positions are more accurate than me!

Maybe this is because I vary rarely work with printed targets and more often building features that are surveyed using scanning and total station so is too easy to put markers in the wrong place...

44

Bug Reports / Cancel build geometry leaves partial depth maps that can be reused

« on: May 30, 2013, 01:19:25 AM »

If you build geometry and cancel before it completes creating depth maps, then returning to the build geometry dialog allows you to reuse those depth maps even though they are incomplete, giving you an incomplete or strangely formed model

If the previous step had not even finished loading photos then you can still 'reuse' depth maps that had not even begun being created and no model is made.

Perhaps this is not a bug and maybe even a desired behaviour or undocumented feature

45

Python and Java API / activeChunk.buildTexture - mapping="Camera"

« on: May 20, 2013, 01:37:27 PM »

Hi there,

This is my first dabble in python programming in photoscan, and my first dabble in python for over 10 years so please excuse my ignorance!

I am just trying individual commands in the console window at present, trying to get it to build texture from a single image.

I have a chunk with 17 aligned cameras and geometry already present (93k faces).

My problem is that having finally worked out how to access cameras and frames, the buildTexture function executes happily but returns false, i.e. the texture is not built.

Does anyone know if I am far off getting this to work?!

Thanks

James

Code: [Select]

>>> import PhotoScan
>>> doc = PhotoScan.app.document
>>> camera = PhotoScan.Cameras(doc.activeChunk)[7]
>>> frame = PhotoScan.Frames(camera)[0]
>>> doc.activeChunk.buildTexture(mapping="camera",width=2048,height=2048,format="U8",camera=frame)
False

Finished processing in 0 sec (exit code 0)
>>>