OpenDroneMap is an open source toolkit for processing aerial drone imagery. Typical drones use simple point-and-shoot cameras, so the images from drones, while from a different perspective, are similar to any pictures taken from point-and-shoot cameras, i.e. non-metric imagery. OpenDroneMap turns those simple images into three dimensional geographic data that can be used in combination with other geographic datasets.
In a word, OpenDroneMap is a toolchain for processing raw civilian UAS imagery to other useful products. What kind of products?
- Point Clouds
- Digital Surface Models
- Textured Digital Surface Models
- Orthorectified Imagery
- Classified Point Clouds (coming soon)
- Digital Elevation Models
- etc.
Open Drone Map now includes state-of-the-art 3D reconstruction work by Michael Waechter, Nils Moehrle, and Michael Goesele. See their publication at http://www.gcc.tu-darmstadt.de/media/gcc/papers/Waechter-2014-LTB.pdf.
The easiest way to run ODM is through Docker. If you don't have it installed,
see the Docker Ubuntu installation tutorial and follow the
instructions through "Create a Docker group". The Docker image workflow
has equivalent procedures for Mac OS X and Windows found at docs.docker.com. Then run the following command which will build a pre-built image and run on images found in $(pwd)/images
(you can change this if you need to, see the wiki for more detailed instructions.
docker run -it --rm -v $(pwd)/images:/code/images -v $(pwd)/odm_orthophoto:/code/odm_orthophoto -v $(pwd)/odm_texturing:/code/odm_texturing opendronemap/opendronemap
** Please note that we need help getting ODM updated to work for 16.10+. Look at #659 or drop into the [gitter][https://gitter.im/OpenDroneMap/OpenDroneMap) for more info.
Download the latest release here
Current version: 0.3.1 (this software is in beta)
- Extract and enter the OpenDroneMap directory
- Run
bash configure.sh install
- Edit the
settings.yaml
file in your favorite text editor. Set theproject-path
value to an empty directory (you will place sub-directories containing individual projects inside). You can add many options to this file, see here - Download a sample dataset from here (about 550MB) and extract it as a subdirectory in your project directory.
- Run
./run.sh odm_data_aukerman
- Enter dataset directory to view results:
- orthophoto: odm_orthophoto/odm_orthophoto.tif
- textured mesh model: odm_texturing/odm_textured_model_geo.obj
- point cloud (georeferenced): odm_georeferencing/odm_georeferenced_model.ply
See below for more detailed installation instructions.
Extract and enter the downloaded OpenDroneMap directory and compile all of the code by executing a single configuration script:
bash configure.sh install
When updating to a newer version of ODM, it is recommended that you run
bash configure.sh reinstall
to ensure all the dependent packages and modules get updated.
For Ubuntu 15.10 users, this will help you get running:
sudo apt-get install python-xmltodict
sudo ln -s /usr/lib/x86_64-linux-gnu/libproj.so.9 /usr/lib/libproj.so
There are some environmental variables that need to be set. Open the ~/.bashrc file on your machine and add the following 3 lines at the end. The file can be opened with gedit ~/.bashrc
if you are using an Ubuntu desktop environment. Be sure to replace the "/your/path/" with the correct path to the location where you extracted OpenDroneMap:
export PYTHONPATH=$PYTHONPATH:/your/path/OpenDroneMap/SuperBuild/install/lib/python2.7/dist-packages
export PYTHONPATH=$PYTHONPATH:/your/path/OpenDroneMap/SuperBuild/src/opensfm
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/your/path/OpenDroneMap/SuperBuild/install/lib
Note that using run.sh
sets these temporarily in the shell.
First you need a set of images, taken from a drone or otherwise. Example data can be obtained from https://github.com/OpenDroneMap/odm_data
Next, you need to edit the settings.yaml
file. The only setting you must edit is the project-path
key. Set this to an empty directory within projects will be saved. There are many options for tuning your project. See the wiki or run python run.py -h
Then run:
python run.py -i /path/to/images project-name
The images will be copied over to the project path so you only need to specify the -i /path/
once. You can also override any variable from settings.yaml here using the command line arguments. If you want to rerun the whole thing, run
python run.py --rerun-all project-name
or
python run.py --rerun-from odm_meshing project-name
The options for rerunning are: 'resize', 'opensfm', 'slam', 'cmvs', 'pmvs', 'odm_meshing', 'mvs_texturing', 'odm_georeferencing', 'odm_orthophoto'
When the process finishes, the results will be organized as follows:
|-- images/
|-- img-1234.jpg
|-- ...
|-- images_resize/
|-- img-1234.jpg
|-- ...
|-- opensfm/
|-- see mapillary/opensfm repository for more info
|-- depthmaps/
|-- merged.ply # Dense Point cloud (not georeferenced)
|-- odm_meshing/
|-- odm_mesh.ply # A 3D mesh
|-- odm_meshing_log.txt # Output of the meshing task. May point out errors.
|-- odm_texturing/
|-- odm_textured_model.obj # Textured mesh
|-- odm_textured_model_geo.obj # Georeferenced textured mesh
|-- texture_N.jpg # Associated textured images used by the model
|-- odm_georeferencing/
|-- odm_georeferenced_model.ply # A georeferenced dense point cloud
|-- odm_georeferenced_model.ply.laz # LAZ format point cloud
|-- odm_georeferenced_model.csv # XYZ format point cloud
|-- odm_georeferencing_log.txt # Georeferencing log
|-- odm_georeferencing_transform.txt# Transform used for georeferencing
|-- odm_georeferencing_utm_log.txt # Log for the extract_utm portion
|-- odm_orthophoto/
|-- odm_orthophoto.png # Orthophoto image (no coordinates)
|-- odm_orthophoto.tif # Orthophoto GeoTiff
|-- odm_orthophoto_log.txt # Log file
|-- gdal_translate_log.txt # Log for georeferencing the png file
Any file ending in .obj or .ply can be opened and viewed in MeshLab or similar software. That includes opensfm/depthmaps/merged.ply
, odm_meshing/odm_mesh.ply
, odm_texturing/odm_textured_model[_geo].obj
, or odm_georeferencing/odm_georeferenced_model.ply
. Below is an example textured mesh:
You can also view the orthophoto GeoTIFF in QGIS or other mapping software:
(Instructions below apply to Ubuntu 14.04, but the Docker image workflow has equivalent procedures for Mac OS X and Windows. See docs.docker.com)
OpenDroneMap is Dockerized, meaning you can use containerization to build and run it without tampering with the configuration of libraries and packages already installed on your machine. Docker software is free to install and use in this context. If you don't have it installed, see the Docker Ubuntu installation tutorial and follow the instructions through "Create a Docker group". Once Docker is installed, the fastest way to use OpenDroneMap is to run a pre-built image by typing:
docker run -it --rm -v $(pwd)/images:/code/images -v $(pwd)/odm_orthophoto:/code/odm_orthophoto -v $(pwd)/odm_texturing:/code/odm_texturing opendronemap/opendronemap
If you want to build your own Docker image from sources, type:
docker build -t my_odm_image .
docker run -it --rm -v $(pwd)/images:/code/images -v $(pwd)/odm_orthophoto:/code/odm_orthophoto -v $(pwd)/odm_texturing:/code/odm_texturing my_odm_image
Using this method, the containerized ODM will process the images in the OpenDroneMap/images directory and output results
to the OpenDroneMap/odm_orthophoto and OpenDroneMap/odm_texturing directories as described in the Viewing Results section.
If you want to view other results outside the Docker image simply add which directories you're interested in to the run command in the same pattern
established above. For example, if you're interested in the dense cloud results generated by PMVS and in the orthophoto,
simply use the following docker run
command after building the image:
docker run -it --rm -v $(pwd)/images:/code/images -v $(pwd)/odm_georeferencing:/code/odm_georeferencing -v $(pwd)/odm_orthophoto:/code/odm_orthophoto my_odm_image
If you want to get all intermediate outputs, run the following command:
docker run -it --rm -v $(pwd)/images:/code/images -v $(pwd)/odm_georeferencing:/code/odm_georeferencing -v $(pwd)/odm_meshing:/code/odm_meshing -v $(pwd)/odm_orthophoto:/code/odm_orthophoto -v $(pwd)/odm_texturing:/code/odm_texturing -v $(pwd)/opensfm:/code/opensfm -v $(pwd)/pmvs:/code/pmvs opendronemap/opendronemap
To pass in custom parameters to the run.py script, simply pass it as arguments to the docker run
command. For example:
docker run -it --rm -v $(pwd)/images:/code/images v $(pwd)/odm_orthophoto:/code/odm_orthophoto -v $(pwd)/odm_texturing:/code/odm_texturing opendronemap/opendronemap --resize-to 1800 --force-ccd 6.16
If you want to pass in custom parameters using the settings.yaml file, you can pass it as a -v volume binding:
docker run -it --rm -v $(pwd)/images:/code/images v $(pwd)/odm_orthophoto:/code/odm_orthophoto -v $(pwd)/odm_texturing:/code/odm_texturing -v $(pwd)/settings.yaml:/code/settings.yaml opendronemap/opendronemap
A web interface and API to OpenDroneMap is currently under active development in the WebODM repository.
Currently we have an experimental feature that uses ORB_SLAM to render a textured mesh from video. It is only supported on Ubuntu 14.04 on machines with X11 support. See the wikifor details on installation and use.
Coming soon...
For documentation, please take a look at our wiki.Check here first if you are having problems. If you still need help, look through the issue queue or create one. There's also a general help chat here.
Help improve our software!
- Try to keep commits clean and simple
- Submit a pull request with detailed changes and test results