Messages

1

General / Re: combine Iphone depth and rgb video data in Agisoft

« on: August 05, 2024, 12:14:31 AM »

Can you take pictures instead of video and try again? Would be much easier. For video you probably need to write some custom converter software and it's very expensive

2

General / Re: combine Iphone depth and rgb video data in Agisoft

« on: August 04, 2024, 04:37:39 PM »

convert rgb and depth to standard images and then import them in metashape

3

General / Re: Correct way to maximise processing speed?

« on: July 22, 2024, 11:35:06 PM »

In my experience the best way is to process valuable data only. For example if you have a 48MP drone, optical resolution may be lower. In this case resizing images to a smaller size may bring the same information and can be much faster to process. Same with out-of-focus area, they don't bring information, so few stacked images are better than many non-stacked images. Some thing for colors.. an object is black and white? Then you don't need colors at all, grayscale will be much faster.

After this there is another level of optimization that involve multiple machine, etc.. but I always prefer to reduce the data first to valuable data only.

4

General / Re: Do you need to "focus stack" your photos if you need a high quality asset?

« on: July 12, 2024, 02:49:47 AM »

For shoes and high quality you definitively need focus stacking so you can use the best lens aperture (ex. f5.6).

My advices:
- Use a rotating table (a manual one is fine too)
- Use a very black background (ex. Musou Black fabric)
- Merge photo with Helicon Focus or a similar software
- Automate Focus Bracketing to avoid unwanted movements

Processing time of all-focused pictures is much faster, because software hasn't to deal with alignment errors.

5

General / Re: focus stacking for photogrammetry

« on: July 12, 2024, 02:45:14 AM »

From my experience there is a big advantage in:
- 3d model details (shooting at f5.6 avoid diffraction errors)
- alignment (no blurry area that leads to alignment issues)
- processing time (aligment is faster because easier with more complete data)

With a Sony camera the process can be very long because there isn't a focus bracketing built-in in many models.
You can use an external app for this, please have a look here

https://www.youtube.com/watch?v=_6fA7urT3VY

for a comparison and a workflow idea.

6

General / Rain simulation

« on: March 13, 2024, 02:37:13 AM »

Hello dear Metashape users,

I was thinking to simulate how water flow on a steep terrain.

Let's say I have the 3D model and the DEM, what's next to do a 3d simulation?

- What software are you using?
- How are you modelling soil water absorption?
- What else are you considering for such simulations?

Thanks all

7

General / Re: 3D Coordinates of image(s) footprint corners added on image EXIF data

« on: March 07, 2024, 12:21:46 AM »

Hi tazzo,

Thanks for your participation, before all.
I mean that every corner of each picture (from drone camera as example) has 3D coordinates when you project these corners on the ground (terrain). For every corner - I mean on each corner for each image: bottom left, bottom right, upper left and upper right corner.

Don't know about that, but I found similar request on this forum but unfortunatelly without right (or outdated) solutions.

Thanks!

J.K.

Be careful because border is often discarded due to distortion compensation, would be more "accurate" to do this on rectified images instead of source images.

8

General / Re: 3D Coordinates of image(s) footprint corners added on image EXIF data

« on: March 06, 2024, 01:00:23 AM »

do you mean geographic coordinates of an orthomosaic image? should be already a geo-tiff

9

General / Re: Can someone make script to create plane from selected 3 markers?

« on: March 06, 2024, 12:55:50 AM »

you don't have to do this, just assign coordinates of your markers and the position can be automatically updated

10

Camera Calibration / Re: A better way to calibrate cameras

« on: February 12, 2024, 03:14:43 AM »

This is done by scale bars, you have two markers with a fixed distance

11

General / Re: Capturing depth data using iPhone LiDAR

« on: February 10, 2024, 03:01:11 AM »

Some people were using it to scan caves where is not practical to bring mirrorless or laserscanner

12

Camera Calibration / Re: A better way to calibrate cameras

« on: February 06, 2024, 02:18:40 AM »

You can't have an absolute fixed calibration because every time will be slightly different. Metashape calculate from images the calibration parameters, chessboard calibration can be useful as a starting point only. Main issue is that focus change the focal lenght, then bringing lens around the world may decenter glass inside and other issues..

13

General / Re: IS IT POSSIBLE TO EXPORT MODEL WITH CAMERAS POSITIONS IN STANDARD 3D FORMATS?!

« on: January 15, 2024, 10:42:28 PM »

FBX is pretty standard because Autodesk is providing an SDK to software developers.

Meshlab support for FBX is fine in my experience, you can try Blender too.

There is a tool from Autodesk called "FBX Review" that's completely free: https://www.autodesk.com/products/fbx/fbx-review
You can try it too.

If your model freeze is because is too big for Meshlab, maybe you can split it in multiple models or decimate it with Metashape before export.

14

General / Re: IS IT POSSIBLE TO EXPORT MODEL WITH CAMERAS POSITIONS IN STANDARD 3D FORMATS?!

« on: January 08, 2024, 02:00:11 PM »

I don't use any Autodesk software. I have MeshLab for some processing / as viewer for FBX files.
FBX is nice because file size is usually much smaller than other formats, and supported by many softwares.
Parameters for the Sketchfab Pro plan are not related to cameras embedded into the 3d model, it's a different thing.
Why you would like to open camera positions? With sketchfab you have more freedom and you can use the view you want.

15

General / Re: scale bar size keeps changing from my input

« on: January 08, 2024, 10:28:59 AM »

when you use the scale bar as input, this is not the only input so will merged with other inputs (ex. points from images) and you may get an error to have the model consistent