Processing large datasets(8000+ images) for 3D model generation using drone imagery in ODM

Hi all, i wanted to process dataset containing around 8000 images for a single 3D model generation covering terrain and forest area with sufficient overlap,double grid missions and at different gimbal tilt as well to capture the treeline more precisely , is there anyway i can increase the cloud upload limit if not now will WebODM Lightning will develop this functionality upon request?. Split and merge option might not help even for offline processing because 3D mesh model and dense point cloud is needed as the end results.

Got this response from Piero when contacted using WebODM contact page
(- We currently don’t have the capacity to process more than 3,000 images per dataset. This might change in the future, as computing power becomes more affordable.
-You could try to spin your own instance of NodeODM on a large machine such as Amazon AWS or Google Cloud Compute?)

Further adding to this, Any particular RAM capacity and Processor you suggest for such datasets so that it won’t run out of memory halfway?. I’ve checked ODM support pages but specific system requirement i couldn’t find for image limits and their usage at certain major steps like mesh and dense cloud generation. Any particular settings suggestion for processing large datasets for 3D model generation in ODM upload options page,this final info alone could save lots of time for me in enabling or modifying certain parameters because i’m not too familiar with photogrammetry requirements for 3D modeling.It would be really helpful if anyone could provide some suggestions.

Documentation has no specific requirements for processing large datasets, but in here you may find useful information. You may want to look at this, I hope it can help.

1 Like

Yes and check this too: What kind of hardware configuration in ClusterODM


Thanks for this @israelbar and @smathermather-cm. Will check it out.


This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.