Can not procress high quality 3D model

[Julia_Oporto]

Hi, I am trying to process a batch of 443 photos that I took with my DJI Phantom 4 RTK drone to create a high quality 3D model.

My computer specs: 32gb ram, 500gb storage with a AMD Ryzen 9 5900HX processor.

Processing options:
auto-boundary: true, crop: 0, debug: true, dem-resolution: 1, feature-quality:ultra, mesh-octree-depth: 12, mesh-size: 600000, min-num-features : 20000, pc-geometric: true, pc-rectify: true, use-3dmesh: true, verbose: true.

I got this error while processing:

[ERROR] Whoops! You ran out of memory! Add more RAM to your computer, if you’re using docker configure it to use more memory, for WSL2 make use of .wslconfig (Advanced settings configuration in WSL | Microsoft Learn), resize your images, lower the quality settings or process the images using a cloud provider (e.g. https://webodm.net).

[Saijin_Naib]

Welcome!

You’ve selected a lot of very memory intensive options.

Try at Defaults first to get a baseline, and let’s work from there.

[Julia_Oporto]

Thanks for the welcome!

I try it with the 3D Model default option:

Processing options: auto-boundary: true; mesh-octree-depth: 12; use-3dmesh: true; pc-quality: high; mesh size: 300.000; resize image: yes - 2048px.
Processing Time: 1.0h:5.0m:43.0s

This is the result

Is there a way to close those gaps in the model?
and improve the quality of the 3d model without using so much ram memory?

[Saijin_Naib]

That mostly depends upon the quality of the input data and survey method.

Can you attach the Report PDF?

[Julia_Oporto]

This is the report

report.pdf (7.7 MB)

[Saijin_Naib]

Maybe try a new preset based off of Default adding/adjusting these parameters:

--feature-quality ultra
--min-num-features 32000
--pc-filter 0
--pc-quality ultra

Some spots on the report show really low counts of being visible in multiple images. It is possible those areas are in really high-contrast/low-detail regions (shadow, specular/highlight), and if so, there isn’t terribly much you can do without trying to pre-process the data, which is tricky.

[Julia_Oporto]

I put to process with those options and I got the same error

[ERROR] Whoops! You ran out of memory! Add more RAM to your computer, if you’re using docker configure it to use more memory, for WSL2 make use of .wslconfig (Advanced settings configuration in WSL | Microsoft Learn), resize your images, lower the quality settings or process the images using a cloud provider (e.g. https://webodm.net).

[Saijin_Naib]

You will need to adjust your WSLConfig as per the log, or possibly more RAM to run at that quality level, then.