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 - indesinenter

Pages: [1]
1
General / Re: Multi-core disabled?
« on: November 27, 2019, 04:38:00 PM »
Also, I update with new information.

In the next process "estimating camera locations" GPU is not working so I guess that is a process do it only by the CPU. In this case I also only have one core working at 100%.

And there isn't a RAM limitation because I'm using just few images (30 images).

How I can take advantage of my i9 and use all cores?

Thanks,

2
General / Multi-core disabled?
« on: November 27, 2019, 04:30:14 PM »
Hi

I am performing a process that takes longer than normal and I have seen that it is only using a processor of my i9. (I attach a picture of this moment)

The workflow is launched through Python.

Have I unconfigured any option? How can I force the use of all processors?

Thanks,

3
Camera Calibration / Re: Marker detection (Treshold size?)
« on: September 22, 2019, 02:15:58 PM »
Hi!

I have been working more on this, and I made a comparison with the same picture (and obviously same relation marker/picture) but with different pixel dimensions. One is 7952 x 5394 (42Mpx) and the other one is 3976 x 2652 (10.5Mpx) (I reduced 50% pixel sizes).

The result is that Marker Detection did not identify any markers in the original image but did identify 20 markers in the reduced image. This leads me to think that there is a maximum internal size (in pixels, not in relation marker/image) for markers that can be detected.

Now that it is easy to have pictures with enough megapixels (more than 24), could it be possible to have access to change this parameter?

And, I suppose that the calibration parameters that I obtain with the reduced image aren't equivalent with the original size? Because the parameters are related with the pixel size of the image?

Thanks!

4
General / Re: Point Filtering (residual values)
« on: September 22, 2019, 01:57:29 PM »
Thanks Alexey for your reply.

And the level parameter is the reprojection error for each point?

I suppose that the elimination of "bad" points (for example, higher than the level: 0.5) increases the quality of the camera calibration? And does the Dense Point Cloud have better geometry/quality?

Thanks

5
General / Point Filtering (residual values)
« on: September 20, 2019, 04:55:04 PM »
Hi!

I've been looking at the forum, but I do not find any similar topic. Is there any option to filter the points for their residual value? It is a parameter that is calculated in the alignment proces and it would be nice be able to eliminate all values ‚Äč‚Äčthat have high residual values. In order to decrease reprojection errors.

For example (I attach a screenshot), I would be interested in eliminating the point that has a very high residual value compared to the others.

Thanks

6
Camera Calibration / Marker detection (Treshold size?)
« on: September 20, 2019, 03:51:15 PM »
Hello,  :)

I am trying to calibrate different cameras with Agisoft Methashape, but I have problems with the automatic detection of markers. Could there be some sort of limiting size / pixels? I find that when the marker (in pixels) is larger than a threshold Methashape does not detect the marker. Can this threshold be modified?

Any help will be appreciated, since I can not make the calibration photos again because the cameras are installed in an inaccessible area.

Attached are two screenshots showing how markers are identified only from a certain distance from the camera. (when the size of the markers is less than a threshold)

Thanks!

7
Python and Java API / IDLE and beginner doubts
« on: February 27, 2019, 07:06:39 PM »
Hi all!

I have two doubts. On the one hand, can Agisoft / Metashape be used with any IDLE external to Agisoft? Or the only interaction with Python has to be from the Agisfot console ?.

I am interested in generating some scripts that make models automatically from photos that I have captured (time-lapse) since 5 months ago and doing it manually one by one is very laborious. Can it be automated by generating a Python code?

The little I have tried to use Scripts that I have been finding have not helped me. I have felt quite useless since I can not even save the file with the doc.save (path) function, it works for me as I put the path I get error.

I appreciate any help.

Thank you

Pages: [1]