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.


Topics - Seboon

Pages: [1]
1
Python and Java API / Get CoordinateSystem
« on: September 09, 2021, 09:19:05 PM »
Hi every one,

Another question, on the command line "Metashape.app.getCoordinateSystem" : locals system provided in GUI disappear (see attachments).
Would it be possible to get them in command line?

Thanks in advance,

Best regards

2
Hi everyone, especially our dear Alexey for whom this message is primarily addressed!


I'm currently working on a custom script that allows user to export orthomosaic at desired scale. It takes the  orthomosaic export dialog of Metashape and add a little option for that.

My issue, if it's really one, is that I don't get the same result for the pixel size as the native dialog of Metashape. It can be seen on the attachments (cap1).

This leads me to some questions regarding  how the software calculate pixel size of an orthomosaic, whether to build or to export.

My orthomosaic has been generated with the same CRS's project ( WGS84/UTM 36N with EGM 2008 , see bottom of cap1 in the console pane)

The command line "Metashape.app.document.chunk.orthomsaic.resolution" matches with my custom dialog but not with the native one. Why??

Thanks to this topic https://www.agisoft.com/forum/index.php?topic=8162.0, the result of the coordinates conversion from meter to degree finally matched the Metashape calculation of pixel size (see cap2)

The CRS I used to generate the orthomosaic is not a geographic system, but a projected one (see cap6.png), so why  I need to convert coordinates from meter to degree to get the same result as the one proposed by the software?

On cap3, I used the native dialog, and switched to the geographic WGS84. When checking the conversion button in meters,  this time it gave me the same result as mine. 

Surely  I'm missing something, but it seems to be inverted, no?
Or does Metashape consider orthomosaic built in geographic mode to use only degree for orthomosaic resolution, only Geographic CRS and not Projected ?

You'll  also find my scipt in attachments ; even it's not yet complete, it can helps!

Thanks in advance for your time and your anwer,

Best regards

3
Bug Reports / Can't capture model view
« on: November 09, 2020, 12:19:32 PM »
Hi,

I can't make  a capture of the model view, either in GUI or within a python script, at very high resolution (more than 100000px...).
Here is the error message I get when trying by gui mode, with the higher resolution proposed by the capture view function:

Code: [Select]
2020-11-09 10:57:52 CaptureView: resolution = 99999x63586, path = C:/Users/Administrateur/Desktop/captureTest.jpg
2020-11-09 10:57:52 renderer_qgl.cpp line 457: glViewport(x, y, width, height) failed: Invalid value

I'm using pro version 1.6.5 build 11249(64bit).

Is there any size limit for the image to be captured?

Any idea please?



4
Python and Java API / Convert event.globalPos() to world coordinates
« on: September 25, 2020, 11:09:01 AM »
Hi Alexey and others,
I'm trying to make a script using a Qrubberband on Orthoview in order to select area on orthomosaic.
I'm able to draw the rubberband and create rectangular shape from it.
The problem is that it doesn't fit the rubberband area.
I think the issue comes from reprojecting the event.globalPos() to world coordinates.
so my question : how to reproject mouseclick coordinates to  world coordinates, as it's done when moving mouse and coordinates shown in the Label  bottom right off the screen??

See my code and cap for more details.

Thanks a lot,

Sébastien

5
Bug Reports / No image data when exporting orthomosaic
« on: September 22, 2020, 03:56:18 PM »
Hey everybody, hey Alexey

When trying  to scale an othomosaic generated with Metashape 1.6.4 in a CAD software, I realized there is no more image data in the properties dialog of the tif.file, as shown in attachment.

The same problem seems to exist on th,e incoming 1.6.5. version.

Could you check on your side if you get the same issue?

By the way, do you plan to add such a feature as mentionned in this topic : https://www.agisoft.com/forum/index.php?topic=6152.msg29792#msg29792

Best regards,

Sébastien


6
Hello,

Could it possible to add the possibility for the user to place the  rotation pivot point of the bounding box by a double-click?

Thanks in advance,

Best regards.

7
Hello,

It would be great if Metashape's Team couls add this feature!

Best regards.

S.Poudroux

8
Python and Java API / Few scripts to enhance Metashape experience
« on: October 30, 2019, 10:43:56 PM »
Hello,

If you are struggling like me in the designing of a python script, you might have restarted Metashape many times, to check the changes made on the script.
Quite frustrating...

So I've written a little script (while struggling...) in order to  reload a custom script without needing to restart the application.

Here's the link : https://github.com/Seboon

You'll find there another script for an auto activation/deactivation of a Node License Key, letting you switch more easily between different computers.

Hope it will feed part of your needs!






9
Bug Reports / Mask from model
« on: June 24, 2019, 09:17:48 AM »
Hello,

Not sure if it' a bug or some mistakes from me, but when using mask from model for all photos in a chunk, I always get an offset on the masked area.
See attachment please.
Any ideas?
Thanks and best regards!

10
Bug Reports / Masking photo bug
« on: June 24, 2019, 08:26:33 AM »
Hello,

I've noticed a bug  in the latest version (1.5.3 build 8432) when using the magic wand tool to mask photos.

When opening  for the first time a photo, there's no problem.
But once the tab is open, if you double-click on another photo to mask, the magic wand won't update, keeping the same mask from the previous photo.

The only way is to turn off the tab photo each time before usinf the tool again.

Best regards,

Sébastien

11
Feature Requests / Back Face Culing in API
« on: March 02, 2018, 11:23:53 AM »
Hello,

It seems that the Back-Face Culling Option  for building Orthomosaic is not implemented in the API?
Could it be possible to add this in next update?
Many thanks!

12
Python and Java API / Reset Transform (Object mode)
« on: January 05, 2018, 04:29:13 PM »
Hello,

Is there anyone here who knows how to reset transormation applied (different from Reset Region) to an object in python scripting?
I've searched in the Python API reference, but found no results.
Thanks!

13
Hello,

For my  work, I oftenly use orthomosaics generated in Photoscan into Qgis, for drawing some contours over it.
And sometimes  I find easier to draw over the shaded or texured mesh only, as i get more details and contrast than I have with orthomosaic.
But the only way - at least, the one I've found - to get this shaded "pattern" into the right position in Qgis is to extract a capture view of the model and then georeference it in Qgis.
I've attached two screenshots for a better understanding:-)

To your mind, would it be possible to get directly a geotiff of the capture view, or even better, to get an option in the export orthomosaic command which would  allow to choose also  between shaded and  textured mesh for the "orthomosaic" generation?

Best regards!

14
Feature Requests / Split the model view window into quad view
« on: April 15, 2017, 10:23:27 AM »
Hello,

It would be very nice if in next releases, we have the possibility to split the model view window into quad view, like in Blender, for example.
Thus, we could inspect more easily the model, with the top, perpective, left and right views displayed on screen at the same time.
Also the possibility to get the real time actualisation on each window when turning/acting on model into only one, and, conversely, an option to lock one of the view.

Do you think such implemantation would be possible?

Many thanks and best regards to your team!


15
General / Gradual selection and optimization
« on: March 12, 2017, 12:32:57 PM »
Hello,

I've seen a lot of topics on the subject, also some tutorials on the web, and i feel disappointed with the gradual selection and optimization steps, as each one contradicts the other one.

My workflow is a summary of what I've found, but not sure it's the best way :

After first alignement, I enter my GCP and activate all of them. Then I optimize cameras, leaving parameters by default.

From here:
1) Reconstruction Uncertainly - Level 10. I do it twice if necessary. Optimization
2) Reprojection Error - Level 0,5. Optimization
3) Number of images - 2. Optimization
4) Projection accuracy - 10% of total points. Optimization

Finally I chek the GCP errors and uncheck those I'm not satisfied with. And run one more optimization.

To your mind, it's a good way?

I'm also wondering if it's better to use all ot the parameters of the gradual selection or not? And if there is an order priority to use them?
And when launch optimization? After each step of gradual selection, or only one time at the end?
Also, is it better to use gradual selection with CGP (after one first optimization)?

Thanks for your help!



Pages: [1]