Messages

1

General / Re: Trouble aligning chunks with cameras

« on: September 29, 2020, 02:45:56 PM »

My bad. Somewhere in the process of merging chunks, the alignment cameras were duplicated in the larger chunk, so when I went to align the smaller chunk, apparently there was ambiguity as to which camera to use (even though the image names were different, e.g. IMG_3675 and IMG_3675.jpg). When I removed the duplicates, it worked fine. Sorry for the bother.

2

General / Trouble aligning chunks with cameras

« on: September 27, 2020, 05:27:54 PM »

Because of the number of cameras (c. 1000) and the limited capacity of my computer, I am building a model of a stone building in chunks. I have successfully aligned chunks of several portions of the structure using camera positions, however I am having difficulty aligning one room interior. To connect the room with the exterior, I included 14 photos from the exterior around the doorway to the room in the room chunk. Alignment etc. went fine, all aligned properly. However, when I then attempt to align this chunk using cameras, nothing happens.

Any ideas where I might be going wrong? Why might the align fail technically?

3

General / Can't move object

« on: April 21, 2019, 03:29:56 PM »

This probably has a simple answer, but sometimes when I am moving and rotating an object and the region, the object button becomes inactive (greyed out) and I can only act on the region. Any ideas?

4

Bug Reports / Re: photos do not display properly when double clicked

« on: June 30, 2018, 10:26:57 PM »

I too am having this trouble and need to inspect my photo masks. I tried the work around by changing the properties of my icon to:

""C:\Program Files\Agisoft\PhotoScan\" --opengl angle"

but still the photo will not display when double-clicked and locks the system, as several other users have reported.

I am running the current version (1.4.2 build 6205). The specifics of my computer are:

Acer Predator G3-571
Intel Core i7-7700HQ CP @ 2.80 GHz 2.81 GHz
RAM: 16.0 GB
64 bit OS
NVIDIA GeForce GTX 1060
OS: Windows 10

Here is the console report when loaded:
2018-06-30 15:22:29 Agisoft PhotoScan Standard Version: 1.4.2 build 6205 (64 bit)
2018-06-30 15:22:29 Platform: Windows
2018-06-30 15:22:29 CPU: Intel(R) Core(TM) i7-7700HQ CPU @ 2.80GHz (laptop)
2018-06-30 15:22:29 CPU family: 6 model: 158 signature: 906E9h
2018-06-30 15:22:29 RAM: 15.9 GB
2018-06-30 15:22:29 OpenGL Vendor: NVIDIA Corporation
2018-06-30 15:22:29 OpenGL Renderer: GeForce GTX 1060/PCIe/SSE2
2018-06-30 15:22:29 OpenGL Version: 4.6.0 NVIDIA 388.16
2018-06-30 15:22:29 Maximum Texture Size: 32768
2018-06-30 15:22:29 Quad Buffered Stereo: not enabled
2018-06-30 15:22:29 ARB_vertex_buffer_object: supported
2018-06-30 15:22:29 ARB_texture_non_power_of_two: supported

Thanks

5

General / Remove GPS information

« on: May 27, 2018, 06:04:42 PM »

Apologies if I have missed a previous post. I wonder if there is a way to ignore the GPS EXIF data in Photoscan without resorting to Python. I have a set of drone photos I collected, the initial few of which have 0s for coordinates, but are otherwise fine. As a result the model produced is aberrant.  I don't need the GPS locations but would like to use all my photos.

Thanks!

6

General / align selected cameras

« on: April 06, 2018, 06:09:50 PM »

I have a general question. I ran a model with 684 photos, going all the way through including creating texture for the mesh model. Took a long time. About 30 of the photos did not align, so I ran align selected cameras, which then aligned about half of the aligned cameras. My question is at what stage am I after doing this? Must I run create dense cloud again (a very long part of the process), or does align selected photos add to the existing cloud, in which case I can go directly to mesh build?

I've asked this before, but why doesn't photoscan do an automatic align check on the unaligned photos after completion of the initial alignment? In other words, why isn't 'align selected photos' performed automatically?

7

Python and Java API / Adjusting console pane information

« on: April 23, 2017, 10:46:54 PM »

Is it possible to programatically turn off the console pane logging information (the time and date to the left)? I would like the console pane to just contain the listing and output.

8

Python and Java API / Missing Index

« on: April 14, 2017, 11:43:02 PM »

The new 1.3.1 manual appears to be missing its module index. Downloaded it twice and only a P on pg. 81.

9

Python and Java API / Re: Extracting profiles

« on: March 23, 2017, 04:39:14 PM »

Kenjy - did you reply to the wrong post?

BTW, I eventually solved this problem by exporting the model to Meshlab, which has provisions for extracting slices/profiles.

10

General / Re: Aligning top and bottom of an object

« on: July 08, 2016, 02:53:25 AM »

I believe markers are a feature of the Pro version only. But applying masks is quite rapid once you have the top and bottom models created. Just strip away the material you don't want first.

11

General / Re: Trouble exporting texture to Meshlab

« on: March 21, 2016, 05:04:43 PM »

As usual, thanks for the excellent user service, Alexey. I had tried both 4096 and 8192 without success. BUT when I just exported it as a 5-letter file name, bingo, it worked. So thanks for that! Hadn't noticed that limitation in Meshlab.

12

General / Trouble exporting texture to Meshlab

« on: March 21, 2016, 04:38:19 PM »

This may be more a Meshlab than Photoscan issue, but several times recently when I export an .obj model from Photoscan, when I import it into Meshlab, the texture will not import. The Meshlab message says:

Error encountered while loading file (xxx.obj).

Error details: Some materials definitions were not found , a default white material is used where no material was available.

The texture .jpg is in fact in the same folder, and I sometimes have models whose texture loads just fine.

As indicated, only the shaded mesh imports. I have tried clicking the UDIM option on and off when exporting, and fixing the topology of the mesh as well. Same result.

Any ideas?

13

Feature Requests / User Interface change requests

« on: March 20, 2016, 04:54:35 PM »

Some minor requests:

1. Ability to "bookmark" a view, to allow the user to return to a particular perspective after making edits. Useful, for instance, to display cut-away images of objects, before and after.

2. Creation of a toolbar for show/hide items, with on-off buttons, saving needless clicks returning to the menu to show/hide multiple items.

3. Ability to drag chunks up and down in the Workspace pane, to better organize them.

4. Ability to choose color for the close hole option. The pink is a little garish sometimes.

5. For batch jobs, have subsequently added tasks use the same chunks as the first by default. That is, if align points uses one chunk out of 7, then build dense cloud should too by default, with the option to modify it. Simple change, but it costs time when many batch jobs are processed.

14

General / Re: Aligning top and bottom of an object

« on: March 09, 2016, 04:20:01 PM »

Thank you all for your replies. I tried James's method and it works great, so now I have a full 3D model with both top and bottom. (BTW, I did not change the background, but will try that tactic for the next go round.)  The workflow blog referred to by Kiesel is also admirably clear, so thanks for that too.

My question still remains, though. How does chunk alignment proceed - sparse or dense point cloud, mesh? Like I mentioned, I removed the supporting surface from both the upright and inverted models in my original attempt, but got that weird alignment. I would have thought having just the point clouds and meshes of the pot itself would have allowed the two chunks to align.

Thanks again for your help.

15

General / Re: Bad imagery alignment

« on: March 09, 2016, 01:12:31 AM »

Do mean align rather than import? If the former, probably because of the varying reflectance patterns of the water from different camera angles.