Split merge for linux noob

I found the documentation on how to perform a split merge and am eager to use it as I only have 64gb of ram. In a linux terminal I pasted:
docker run -ti --rm -v /my/project:/datasets/code opendronemap/odm --project-path /datasets --split 400 --split-overlap 100
As my project has 1103 images, I adapted it to:
sudo docker run -ti --rm -v /mnt/f46c40e7-907c-4634-9444-f849305741f9/AirDrone:/backfield/code opendronemap/odm --project-path /datasets --split 300 --split-overlap 80
The issue that I’m having, is do I keep the colon after “AirDrone”(the folder holding the images) and why is “/code” included? Is this a necessary placeholder?

Thanks.

2 Likes

Can anyone explain which portions of this string needs “customized” and how to customize it? Or… where I can find this information? Google and youtube are incredibly sparse with information on this.

docker run -ti --rm -v /my/project:/datasets/code opendronemap/odm --project-path /datasets --split 400 --split-overlap 100

Thanks in advance!

1 Like

Welcome!

What is the actual path to your images?

1 Like

Hi Saijin!

My path to my images for say… the backfield… is:

/mnt/f46c40e7-907c-4634-9444-f849305741f9/AirDrone/backfield

edit: I’m not sure if this is important, but this folder… /backfield/ only contains jpeg images. It contains no subfolders or other files.

1 Like

Hopefully this helps:

I have the brighton_beach dataset from ODMData in my home directory, like so:

/home/username/Sources/ODMData/brighton_beach

Inside that folder, there is the images directory that contains all the photos I’ll process.

My Docker run command looks like this:

docker run -ti --rm -v /home/username/Sources/ODMData/brighton_beach:/datasets/brighton_beach opendronemap/odm:latest --project-path /datasets brighton_beach

In your case, it should probably look more like this:

docker run -ti --rm -v /mnt/f46c40e7-907c-4634-9444-f849305741f9/AirDrone/backfield:/datasets/backfield opendronemap/odm:latest --project-path /datasets backfield

This took me a bit, so I’ve filed a note to improve the Docs and examples around this workflow.

3 Likes

Thanks Saijin! I did some work with what Saijin gave me.
The name of my path is this:

/mnt/f46c40e7-907c-4634-9444-f849305741f9/AirDrone/backfield

I’m not exactly sure why, but the script looks for a subfolder at:

/mnt/f46c40e7-907c-4634-9444-f849305741f9/AirDrone/backfield/images

I made a subfolder called “images” in my backfield folder. I put all of my images in this subfolder.

I added the flags based on how much I though my ram could handle

--split 300 --split-overlap 80

Then I ran it. If you aren’t sure how to run webodm from command line, I used these three commands:

cd WebODM
./webodm.sh start

Finally, I ran the string to start my customized script:

sudo docker run -ti --rm -v /mnt/f46c40e7-907c-4634-9444-f849305741f9/AirDrone/backfield:/datasets/backfield opendronemap/odm:latest --project-path /datasets backfield --split 300 --split-overlap 80

I hope that this helps someone!

Thanks!

2 Likes

Yep! Me at least when I go to draw up the documentation update :rofl: Thanks for your contribution!

2 Likes

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