Skip to content

OpenDroneMap (ODM)

OpenDroneMap (ODM) is an open source command line toolkit for processing aerial images. It is used for instance in making DEMs, DSMs, point clouds and orthophotos photogrammetrically.

Available

OpenDroneMap is available in Puhti with following versions:

  • 3.0.4
  • 2.8.8

Usage

OpenDroneMap is available in Puhti as a Singularity container

To run OpenDroneMap: 1) Copy your aerial images to Puhti. OpenDroneMap requires the folder names to end with code/images, for example /scratch/project_2000XXX/odm/code/images. 2) Write a batch job script (see below) 3) Submit your OpenDroneMap batch job.

  • According to our tests, a project with ~300 images is optimal to run with 8-12 CPU cores; adjust the number of CPUs on the --cpus-per-task line. OpenDroneMap can use only one computing node in processing which means that it is limited to a maximum of 40 CPU cores per job.
  • --project-path - the place where images are stored, without the code/images part.
  • --max-concurrency - the number of threads used in several steps of ODM processing, here set to same number as reserved cores.
  • It is possible to add additional arguments to the end of the command.
  • If your images cover a very large area, see ODM Splitting Large Datasets documentation Below you can find an example batch job script. Adjust --account, --cpus-per-task, --time and --mem-per-cpu to your needs. 
    #!/bin/bash
#SBATCH --account=<YOUR-CSC-PROJECT>
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=8
#SBATCH --partition=small
#SBATCH --time=02:00:00
#SBATCH --mem-per-cpu=3G

module load opendronemap
apptainer_wrapper run --project-path /scratch/project_2000XXX/odm --max-concurrency $SLURM_CPUS_PER_TASK

3) Outputs are available in code-folder, for example /scratch/project_2000XXX/odm/code

OpenDroneMap with compute node's local NMVE disk

OpenDroneMap reads and writes a lot to disk, so running it is slightly (~15%) faster using compute node's local NMVE disk. Below is example file for using OpenDroneMap with NMVE disk.

#!/bin/bash
#SBATCH --account=<YOUR-CSC-PROJECT>
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=8
#SBATCH --partition=small
#SBATCH --time=02:00:00
#SBATCH --mem-per-cpu=3G
#SBATCH --gres=nvme:30

#ODM project folder, that has code/images folder.
odm_dir=/scratch/project_2000599/odm/

echo "Copying input images from Puhti scratch to compute node local disk"
rsync -r $odm_dir/code $LOCAL_SCRATCH

module load opendronemap
apptainer_wrapper run --project-path $LOCAL_SCRATCH --max-concurrency $SLURM_CPUS_PER_TASK

echo "Copying outputs from Puhti scratch to compute node local disk"
rsync -r $LOCAL_SCRATCH/* $odm_dir

License

OpenDroneMap is distributed under the GNU General Public License (GPL) version 3. Full OpenDroneMap license

Citation

OpenDroneMap Authors, ODM - A command line toolkit to generate maps, point clouds, 3D models and DEMs from drone, balloon or kite images. OpenDroneMap/ODM GitHub Page 2020; https://github.com/OpenDroneMap/ODM

Acknowledgement

Please acknowledge CSC and Geoportti in your publications, it is important for project continuation and funding reports. As an example, you can write "The authors wish to thank CSC - IT Center for Science, Finland (urn:nbn:fi:research-infras-2016072531) and the Open Geospatial Information Infrastructure for Research (Geoportti, urn:nbn:fi:research-infras-2016072513) for computational resources and support".

Installation

OpenDroneMap was installed to Puhti with Apptrainer using the OpenDroneMap Docker image from Dockerhub provided by OpenDroneMap community. OpenDroneMap commands have not been wrapped with Tykky, because mapping of folders is always needed at run-time.

References