I create a docker image to help AWS users make orthomosaics using S3 and ODM. You can access it at Docker Hub.
I have some limitations like, I can only make orthomosaics with up to 450 images (in 32Gb and 8 cores RAM). I have problems with processing similar photos (I’m solving this). I’m trying to process larger orthomosaics. Do you have some tips?
Thanks for your attention.
Welcome!
Really neat Docker image!
As to your questions, what processing options have you set as defaults in your Docker image? Also, is it pre-configured to use any SWAP or virtual memory alongside the 32GB RAM you’ve chosen as the VM size?
Could you explain a bit more about the problems with similar image processing? Maybe some screenshots or posting the datasets in DroneDB.app would help.
Larger projects might require more RAM and SWAP, and/or lower processing parameters.
I’m still not using SWAP.
I was using these parameters (–resize-to 1024 --depthmap-resolution 160) and it performed well, up to 450 images in under 30 minutes.
Now I’m trying to use (–feature-type hahog --split 64) to fix these errors with similar photos. I see it takes longer, but it works.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.