A command line toolkit to generate maps, point clouds, 3D models and DEMs from drone, balloon or kite images. 📷

Overview

ODM Logo

An open source command line toolkit for processing aerial drone imagery. ODM turns simple 2D images into:

  • Classified Point Clouds
  • 3D Textured Models
  • Georeferenced Orthorectified Imagery
  • Georeferenced Digital Elevation Models

images-diag

The application is available for Windows, Mac and Linux and it works from the command line, making it ideal for power users, scripts and for integration with other software.

If you would rather not type commands in a shell and are looking for a friendly user interface, check out WebODM.

Quickstart

The easiest way to run ODM is via docker. To install docker, see docs.docker.com. Once you have docker installed and working, you can run ODM by placing some images (JPEGs or TIFFs) in a folder named “images” (for example C:\Users\youruser\datasets\project\images or /home/youruser/datasets/project/images) and simply run from a Command Prompt / Terminal:

# Windows
docker run -ti --rm -v c:/Users/youruser/datasets:/datasets opendronemap/odm --project-path /datasets project

# Mac/Linux
docker run -ti --rm -v /home/youruser/datasets:/datasets opendronemap/odm --project-path /datasets project

You can pass additional parameters by appending them to the command:

docker run -ti --rm -v /datasets:/datasets opendronemap/odm --project-path /datasets project [--additional --parameters --here]

For example, to generate a DSM (--dsm) and increase the orthophoto resolution (--orthophoto-resolution 2) :

docker run -ti --rm -v /datasets:/datasets opendronemap/odm --project-path /datasets project --dsm --orthophoto-resolution 2

Viewing Results

When the process finishes, the results will be organized as follows:

|-- images/
    |-- img-1234.jpg
    |-- ...
|-- opensfm/
    |-- see mapillary/opensfm repository for more info
|-- odm_meshing/
    |-- odm_mesh.ply                    # A 3D mesh
|-- odm_texturing/
    |-- odm_textured_model.obj          # Textured mesh
    |-- odm_textured_model_geo.obj      # Georeferenced textured mesh
|-- odm_georeferencing/
    |-- odm_georeferenced_model.laz     # LAZ format point cloud
|-- odm_orthophoto/
    |-- odm_orthophoto.tif              # Orthophoto GeoTiff

You can use the following free and open source software to open the files generated in ODM:

  • .tif (GeoTIFF): QGIS
  • .laz (Compressed LAS): CloudCompare
  • .obj (Wavefront OBJ), .ply (Stanford Triangle Format): MeshLab

Note! Opening the .tif files generated by ODM in programs such as Photoshop or GIMP might not work (they are GeoTIFFs, not plain TIFFs). Use QGIS instead.

API

ODM can be made accessible from a network via NodeODM.

Documentation

See http://docs.opendronemap.org for tutorials and more guides.

Forum

We have a vibrant community forum. You can search it for issues you might be having with ODM and you can post questions there. We encourage users of ODM to partecipate in the forum and to engage with fellow drone mapping users.

Snap Package

ODM is now available as a Snap Package from the Snap Store. To install you may use the Snap Store (available itself as a Snap Package) or the command line:

sudo snap install opendronemap

To run, you will need a terminal window into which you can type:

opendronemap

# or

snap run opendronemap

# or

/snap/bin/opendronemap

Snap packages will be kept up-to-date automatically, so you don't need to update ODM manually.

GPU Acceleration

ODM has support for doing SIFT feature extraction on a GPU, which is about 2x faster than the CPU on a typical consumer laptop. To use this feature, you need to use the opendronemap/odm:gpu docker image instead of opendronemap/odm and you need to pass the --gpus all flag:

docker run -ti --rm -v c:/Users/youruser/datasets:/datasets --gpus all opendronemap/odm:gpu --project-path /datasets project

When you run ODM, if the GPU is recognized, in the first few lines of output you should see:

[INFO]    Writing exif overrides
[INFO]    Maximum photo dimensions: 4000px
[INFO]    Found GPU device: Intel(R) OpenCL HD Graphics
[INFO]    Using GPU for extracting SIFT features

The SIFT GPU implementation is OpenCL-based, so should work with most graphics card (not just NVIDIA).

If you have an NVIDIA card, you can test that docker is recognizing the GPU by running:

docker run --rm --gpus all nvidia/cuda:10.0-base nvidia-smi

If you see an output that looks like this:

Fri Jul 24 18:51:55 2020       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 440.82       Driver Version: 440.82       CUDA Version: 10.2     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |

You're in good shape!

See https://github.com/NVIDIA/nvidia-docker and https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html#docker for information on docker/NVIDIA setup.

WSL or WSL2 Install

Note: This requires that you have installed WSL already by following the instructions on Microsoft's Website.

You can run ODM via WSL or WSL2 by downloading the rootfs.tar.gz file from the releases page on GitHub. Once you have the file saved to your Downloads folder in Windows, open a PowerShell or CMD window by right-clicking the Flag Menu (bottom left by default) and selecting "Windows PowerShell", or alternatively by using the Windows Terminal from the Windows Store.

Inside a PowerShell window, or Windows Terminal running PowerShell, type the following:

# PowerShell
wsl.exe --import ODM $env:APPDATA\ODM C:\path\to\your\Downloads\rootfs.tar.gz

Alternatively if you're using CMD.exe or the CMD support in Windows Terminal type:

# CMD
wsl.exe --import ODM %APPDATA%\ODM C:\path\to\your\Downloads\rootfs.tar.gz

In either case, make sure you replace C:\path\to\your\Downloads\rootfs.tar.gz with the actual path to your rootfs.tar.gz file.

This will save a new Hard Disk image to your Windows AppData folder at C:\Users\username\AppData\roaming\ODM (where username is your Username in Windows), and will set-up a new WSL "distro" called ODM.

You may start the ODM distro by using the relevant option in the Windows Terminal (from the Windows Store) or by executing wsl.exe -d ODM in a PowerShell or CMD window.

ODM is installed to the distro's /code directory. You may execute it with:

/code/run.sh

Updating ODM in WSL

The easiest way to update the installation of ODM is to download the new rootfs.tar.gz file and import it as another distro. You may then unregister the original instance the same way you delete ODM from WSL (see next heading).

Deleting an ODM in WSL instance

wsl.exe --unregister ODM

Finally you'll want to delete the files by using your Windows File Manager (Explorer) to navigate to %APPDATA%, find the ODM directory, and delete it by dragging it to the recycle bin. To permanently delete it empty the recycle bin.

If you have installed to a different directory by changing the --import command you ran to install you must use that directory name to delete the correct files. This is likely the case if you have multiple ODM installations or are updating an already-installed installation.

Native Install (Ubuntu 20.04)

You can run ODM natively on Ubuntu 20.04 LTS (although we don't recommend it):

  1. Download the source from here
  2. Run bash configure.sh install
  3. Download a sample dataset from here (about 550MB) and extract it in /datasets/aukerman
  4. Run ./run.sh --project-path /datasets odm_data_aukerman

Updating a native installation

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.

Build From Source

If you want to rebuild your own docker image (if you have changed the source code, for example), from the ODM folder you can type:

docker build -t my_odm_image --no-cache .

When building your own Docker image, if image size is of importance to you, you should use the --squash flag, like so:

docker build --squash -t my_odm_image .

This will clean up intermediate steps in the Docker build process, resulting in a significantly smaller image (about half the size).

Experimental flags need to be enabled in Docker to use the --squash flag. To enable this, insert the following into the file /etc/docker/daemon.json:

{
   "experimental": true
}

After this, you must restart docker.

Developers

Help improve our software! We welcome contributions from everyone, whether to add new features, improve speed, fix existing bugs or add support for more cameras. Check our code of conduct, the contributing guidelines and how decisions are made.

For Linux users, the easiest way to modify the software is to make sure docker is installed, clone the repository and then run from a shell:

$ DATA=/path/to/datasets ./start-dev-env.sh

Where /path/to/datasets is a directory where you can place test datasets (it can also point to an empty directory if you don't have test datasets).

Run configure to set up the required third party libraries:

(odmdev) [user:/code] master+* ± bash configure.sh reinstall

You can now make changes to the ODM source. When you are ready to test the changes you can simply invoke:

(odmdev) [user:/code] master+* ± ./run.sh --project-path /datasets mydataset

If you have questions, join the developer's chat at https://community.opendronemap.org/c/developers-chat/21

  1. Try to keep commits clean and simple
  2. Submit a pull request with detailed changes and test results
  3. Have fun!

Credits

ODM makes use of several libraries and other awesome open source projects to perform its tasks. Among them we'd like to highlight:

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

Issues
  • Segfault while merging depthmaps

    Segfault while merging depthmaps

    Hi, I'm running the latest ODM Docker image (pulled last night) on a VPS running Debian with 24GB of RAM and 8 CPU cores. I'm currently trying to stitch together a survey that consists of around 270 photos, totaling 1-1.5GB. During processing, I get this:

    2017-03-10 13:48:42,183 Cleaning depthmap for image DJI_0062.JPG 2017-03-10 13:48:43,404 Cleaning depthmap for image DJI_0021.JPG 2017-03-10 13:48:44,682 Cleaning depthmap for image DJI_0060.JPG 2017-03-10 13:48:46,072 Cleaning depthmap for image DJI_0162.JPG 2017-03-10 13:48:47,415 Cleaning depthmap for image DJI_0236.JPG 2017-03-10 13:48:52,428 Cleaning depthmap for image DJI_0050.JPG 2017-03-10 13:48:53,772 Cleaning depthmap for image DJI_0052.JPG 2017-03-10 13:48:55,157 Cleaning depthmap for image DJI_0228.JPG 2017-03-10 13:48:56,464 Cleaning depthmap for image DJI_0087.JPG 2017-03-10 13:48:57,821 Cleaning depthmap for image DJI_0124.JPG 2017-03-10 13:48:59,103 Cleaning depthmap for image DJI_0118.JPG 2017-03-10 13:49:00,459 Cleaning depthmap for image DJI_0079.JPG 2017-03-10 13:49:01,845 Cleaning depthmap for image DJI_0241.JPG 2017-03-10 13:49:06,852 Cleaning depthmap for image DJI_0051.JPG 2017-03-10 13:49:08,309 Merging depthmaps Segmentation fault (core dumped) Traceback (most recent call last): File "/code/run.py", line 55, in plasm.execute(niter=1) File "/code/scripts/opensfm.py", line 90, in process (context.pyopencv_path, context.opensfm_path, tree.opensfm)) File "/code/opendm/system.py", line 28, in run raise Exception("Child returned {}".format(retcode)) Exception: Child returned 139 Any ideas? I'm fairly new to the project and I'm not sure where to dig in.

    troubleshooting 
    opened by bstempi 63
  • Align of resulting mosaic (in case of multispectral imaging)

    Align of resulting mosaic (in case of multispectral imaging)

    Each band of multispectral sensors (e.g. Micasense Rededge, Sequoia) is taken with separate lens, where different angle for each band is appied (each band image is in separate file). When sources images are merged to multiband image, there bands are significantly shifted each other. Therefore, each band must be processed separately in case of mosaicking. Unfortunately, when merging resulting mosaics, similar shift is present: odm Possibility of aligning resulting mosaics would be appreciated.

    Thanks.

    enhancement 
    opened by ivopavlik 59
  • Update OpenSfM

    Update OpenSfM

    This PR brings in the latest changes to OpenSfM, which among other things include the reprojection of derivatives which are now computed analytically instead of autodiff-ed (20%-45% speed-up on opensfm reconstruct)

    WIP

    enhancement 
    opened by pierotofy 45
  • Planar Reconstruction

    Planar Reconstruction

    This PR adds support for really fast planar reconstructions (e.g. agricultural fields)

    Requirements:

    • Constant flight altitude
    • Nadir-only camera shots
    • Single pass lawnmower pattern
    • Planar or mostly planar terrain
    • Single camera (no multi-camera support at this moment). Multispectral images are fine.

    Then to obtain an orthophoto really quickly, one can pass:

    --sfm-algorithm planar --matcher-neighbors 4 --fast-orthophoto

    If one needs a full 3D model from the mostly planar scene, one can omit the --fast-orthophoto flag and a full 3D reconstruction will still take place.

    Experimental! :boom:

    opened by pierotofy 44
  • Fewer than 3 GCPs have correspondences in the generated model.

    Fewer than 3 GCPs have correspondences in the generated model.

    I have been repeatedly encountering an issue with the GCPs by running ODM with the Toledo input and also with my own data. I have checked the GPS locations in the exif data from the images. The GCPs are in the right locations and the coordinates are in UTM with the correct header. But I get always the same error:

    Error in Georef: Fewer than 3 GCPs have correspondences in the generated model. [ERROR] Georeferencing failed. Traceback (most recent call last): File "/home/vman/vman/OpenDroneMap/scripts/odm_georeferencing.py", line 124, in process '-outputCoordFile {coords}'.format(**kwargs)) File "/home/vman/vman/OpenDroneMap/opendm/system.py", line 28, in run raise Exception("Child returned {}".format(retcode)) Exception: Child returned 1

    The odm_georeferencing_log.txt files says at the bottom: Successfully loaded /home/vman/vman/OpenDroneMap/tests/test_data/odm_texturing/odm_textured_model.obj. Error in Georef: Fewer than 3 GCPs have correspondences in the generated model.

    I have more than 3 GCPs in the gcp_list.txt file. Everything works fine by ignoring the gcp_list.txt file (--use-exif option).

    Is there still something I am overseeing?

    Thanks Volker

    bug 
    opened by vroeb 43
  • Fast Orthophoto via sparse reconstruction

    Fast Orthophoto via sparse reconstruction

    This PR adds support for a new --fast-orthophoto flag, which skips the opensfm dense reconstruction step and uses the sparse reconstruction to generate an orthophoto, while skipping the generation of the full 3D model (the poisson mesh). This is going to benefit users that just want an orthophoto and works especially well on flat areas such as farm fields.

    List of changes:

    • Added --fast-orthophoto, --mesh-neighbors and --mesh-resolution parameters.
    • Removed old 2.5D mesh module and related flags.
    • Wrote new 2.5D mesh program, based on VTK7. We cannot use the binaries that ship with Ubuntu because the parallel features, which are critical for good performance, are not available in the binary packages. VTK7 adds time to the compilation, but doesn't increase image size in docker. The 2.5D mesh works by interpolating a DSM using shepard's method and performing a greedy terrain decimation, which allows us to retain the shape of buildings even with few points and fill-in gaps in the terrain. It's multithreaded, so meshing will be relatively fast especially on machines with multiple cores.

    image

    image

    Fast orthophoto:

    image

    • Cleaned up code that wasn't being used anymore
    • Made learner docker images, most of the files in /code/SuperBuild/src are not needed because they are installed in /code/SuperBuild/install (with a few exceptions)
    • Users can continue computing a DSM/DTM from the sparse point cloud when --fast-orthophoto is used.
    • Increased the min-num-features parameter default to 8000. This is both to help users get better orthophotos using the --fast-orthophoto flag, but also to help users get better overall results even when using the dense reconstruction. Often times 4000 isn't sufficient for areas of high vegetation and new users seem confused. If this is not a good idea, please let me know.
    • Users can keep using the dense reconstruction for generating the orthophoto and they can choose to use the new 2.5D mesh module. The --use-25dmesh flag remains. When --fast-orthophoto is used, --use-25dmesh is implicitly set.

    I recommend testing also with the changes introduced here: https://github.com/OpenDroneMap/mvs-texturing/pull/1

    The PR is stable enough to be tested, but do not merge yet. I'd love to hear feedback and do some more testing, especially to make sure that nothing in the existing pipeline broke. 😄

    enhancement 
    opened by pierotofy 42
  • Does ODM support Tif format

    Does ODM support Tif format

    I often use the multi-spectral camera to capture images in the experiment filed. So, I wonder if I can use Tif format files with multi-band, e.g. 10 bands in one Tif image, as input source images after some modifications. Is it a heavy work if I do this modifications? Thank you.

    enhancement help wanted 
    opened by xialang2012 42
  • Add SLAM module

    Add SLAM module

    This PR adds a SLAM module to OpenDroneMap so that it is possible to use videos as input instead of still images. It uses the open-source library ORB_SLAM2 to compute the camera trajectory and then continues with PMVS and the rest of the pipeline.

    Building the SLAM module is set as an option in the CMake files and the default is not to build it.

    Here's a short guide on how to use it

    opened by paulinus 41
  • Generating a footprint and average % overlap

    Generating a footprint and average % overlap

    At some point during reconstruction, is it possible to produce an estimate of overlap or generate something like this:

    footprint

    I think it would help with benchmarking and being able to discriminate between bad runs and bad photosets.

    enhancement help wanted 
    opened by dakotabenjamin 41
  • OpenCV-3.3.1 w/ nvidia/cuda:8.0-devel-ubuntu16.04 image for GPU processing tests

    OpenCV-3.3.1 w/ nvidia/cuda:8.0-devel-ubuntu16.04 image for GPU processing tests

    OpenCV-2.4.xx w/ Cuda does not recognize newer classes of NVidia GPU when building, aka only up to Kepler. I need Pascal, therefore need to build with OpenCV-3.3.0

    Updated SuperBuild/CMakeLists.txt & Extenal-xxxx.cmake to newest versions: Ecto --> 0.6.12 OpenCV-2.4.11 --> 3.3.0 PCL 1.8.0 -->1.8.1

    On host: $ docker build --tag gei/opendronemap:cuda-8.0-opencv-3.3.0 Builds successfully through [rest of Dockerfile is commented out]: $ cd SuperBuild/build && cmake .. && make -j20 no errors

    $ docker run -it --name odm-cuda-8.0-opencv-3.3.0 gei/opendronemap:cuda-8.0-opencv-3.3.0 bash In container -> moving up to next Dockerfile step: $ cd ../.. && mkdir build && cd build && cmake .. a few warnings, but -- Configuring done -- Generating done -- Build files have been written to: /code/build

    $ make -j20 throws errors at odm_texturing: [ 86%] Linking CXX executable ../../bin/odm_meshing [ 86%] Built target odm_meshing [ 89%] Linking CXX executable ../../bin/odm_texturing CMakeFiles/odm_texturing.dir/src/OdmTexturing.cpp.o: In function OdmTexturing::loadCameras()': OdmTexturing.cpp:(.text+0x2a3b): undefined reference tocv::imread(cv::String const&, int)' CMakeFiles/odm_texturing.dir/src/OdmTexturing.cpp.o: In function OdmTexturing::createTextures()': OdmTexturing.cpp:(.text+0x84ef): undefined reference tocv::imread(cv::String const&, int)' OdmTexturing.cpp:(.text+0x8a43): undefined reference to cv::imwrite(cv::String const&, cv::_InputArray const&, std::vector<int, std::allocator<int> > const&)' OdmTexturing.cpp:(.text+0x8d35): undefined reference tocv::imwrite(cv::String const&, cv::_InputArray const&, std::vector<int, std::allocator > const&)' collect2: error: ld returned 1 exit status modules/odm_texturing/CMakeFiles/odm_texturing.dir/build.make:479: recipe for target 'bin/odm_texturing' failed make[2]: *** [bin/odm_texturing] Error 1 CMakeFiles/Makefile2:338: recipe for target 'modules/odm_texturing/CMakeFiles/odm_texturing.dir/all' failed make[1]: *** [modules/odm_texturing/CMakeFiles/odm_texturing.dir/all] Error 2 make[1]: *** Waiting for unfinished jobs.... [ 93%] Linking CXX executable ../../bin/odm_orthophoto CMakeFiles/odm_orthophoto.dir/src/OdmOrthoPhoto.cpp.o: In function OdmOrthoPhoto::createOrthoPhoto()': OdmOrthoPhoto.cpp:(.text+0x916c): undefined reference tocv::imread(cv::String const&, int)' OdmOrthoPhoto.cpp:(.text+0x96bb): undefined reference to cv::imwrite(cv::String const&, cv::_InputArray const&, std::vector<int, std::allocator<int> > const&)' collect2: error: ld returned 1 exit status modules/odm_orthophoto/CMakeFiles/odm_orthophoto.dir/build.make:453: recipe for target 'bin/odm_orthophoto' failed make[2]: *** [bin/odm_orthophoto] Error 1 CMakeFiles/Makefile2:283: recipe for target 'modules/odm_orthophoto/CMakeFiles/odm_orthophoto.dir/all' failed make[1]: *** [modules/odm_orthophoto/CMakeFiles/odm_orthophoto.dir/all] Error 2 [ 96%] Linking CXX executable ../../bin/odm_25dmeshing [ 96%] Built target odm_25dmeshing [100%] Linking CXX executable ../../bin/odm_georef CMakeFiles/odm_georef.dir/src/Georef.cpp.o: In functionGeoref::performGeoreferencingWithGCP()': Georef.cpp:(.text+0xaec2): undefined reference to `cv::imread(cv::String const&, int)' collect2: error: ld returned 1 exit status modules/odm_georef/CMakeFiles/odm_georef.dir/build.make:506: recipe for target 'bin/odm_georef' failed make[2]: *** [bin/odm_georef] Error 1 CMakeFiles/Makefile2:173: recipe for target 'modules/odm_georef/CMakeFiles/odm_georef.dir/all' failed make[1]: *** [modules/odm_georef/CMakeFiles/odm_georef.dir/all] Error 2 Makefile:83: recipe for target 'all' failed make: *** [all] Error 2

    Ideas? I'm thinking that possibly the ODM scripts calls need to be tuned for the newer versions of software?

    opened by PeterSprague 38
  • Python-port: odm_georeferencing, odm_orthophoto

    Python-port: odm_georeferencing, odm_orthophoto

    Please assist...

    ...run.py with --use-opensfm True and --odm_georeferencingGcp False produces no orthophoto. The arguments do produce .ply meshes and texturing with mosaicked photos on top though? odm_georeferencing folder under results also then contain no files. Also ran ./install.sh recently through python-port.

    Terminal output shows that I must specify ...Gcp False even though I did include the argument.

    Run from Uhuntu 14

    opened by rion-saeon 38
  • Add support for P4 camera+sun calibration

    Add support for P4 camera+sun calibration

    Currently the Phantom 4 multispectral sun sensor calibration does not work properly. Some modifications to the code need to be implemented to support it.

    Reference: https://dl.djicdn.com/downloads/p4-multispectral/20200717/P4_Multispectral_Image_Processing_Guide_EN.pdf

    Community thread: https://community.opendronemap.org/t/dji-p4m-segmented-orthophoto/12039/15

    bug enhancement help wanted 
    opened by pierotofy 0
  • Docker Image build fails

    Docker Image build fails

    Running docker build -t [tags] . using an unmodified clone of ODM (https://github.com/OpenDroneMap/ODM) failed setting up GDAL with pip.

    Building a very slightly modified ODM the same way resulted in a lot of errors, while the docker pull... command works just fine.

    Hardware: Apple : M1 MacBook Air, 8gb ram, 512gb SSD Windows11 : Intel i5 7th gen 7300HQ, NVIDIA GTX 1050 4gb, 16gb Ram (Laptop), 1tb Storage(non SSD) Ubuntu : Running Virtual machine on Mac M1

    I have tried running the clone on all the above machines.

    This issue can be reproduced while building an Unmodified clone of the ODM using the DockerFile provided using the docker build -t . The issue every time is a Different one, I will try uploading the logs later on. Sometimes it is an OpenCv2 error or a PCL error, Make Build Failed.

    possible bug 
    opened by NightRaven-spec 3
  • add dmpdtm.py

    add dmpdtm.py

    Pull request with the intent of integrating some of the magic of DMPdtm from Josh @vonnonn's adaptation of Brad Chambers implementation of Mungus et al., 2014 https://github.com/vonnonn/poodles/blob/main/DMPdtm.py

    opened by smathermather 4
  • Failed to build/install nativly on ubuntu 20.04

    Failed to build/install nativly on ubuntu 20.04

    I've download the code and activated a dedicated python environment (3.8)

    $ bash configure.sh install
    <lots of printouts>
    [ 88%] Performing install step for 'poissonrecon'
    [ 89%] Completed 'poissonrecon'
    [ 89%] Built target poissonrecon
    make: *** [Makefile:84: all] Error 2
    

    I've added set(CMAKE_VERBOSE_MAKEFILE ON) to CMakeLists.txt, but no helpful information is provided. GNU Make 4.2.1 gcc (Ubuntu 9.4.0-1ubuntu1~20.04) 9.4.0

    help wanted possible bug 
    opened by assafge 1
  • GPU processing crashes and fallsback to CPU processing

    GPU processing crashes and fallsback to CPU processing

    How did you install OpenDroneMap? (Docker, natively, ...)?

    Docker GPU image

    What's your browser and operating system?

    Ubuntu 20.04

    What is the problem?

    GPU processing crashes and falls back to CPU processing during the feature extraction leaving this message:

    /code/SuperBuild/install/bin/opensfm/bin/opensfm: line 12: 65 Segmentation fault (core dumped) "$PYTHON" "$DIR"/opensfm_main.py "[email protected]" [WARNING] GPU SIFT extraction failed, maybe the graphics card is not supported? Attempting fallback to CPU

    How can we reproduce this? (What steps did you do to trigger the problem? What parameters are you using for processing? If possible please include a copy of your dataset uploaded on Google Drive or Dropbox. Be detailed)

    I'm using the following command:

    docker run -ti --rm --gpus all -v /home/manu/project/:/workspace opendronemap/odm:gpu --project-path /workspace datasets --dsm --dtm --min-num-features 15000 --matcher-neighbors 20 --dem-resolution 2.0 --feature-quality high --rerun-all

    I'm running the docker in an nvidia rtx 3080, with driver version 510.47.03 The memory of the card is nowhere near full (normally using 2/10 GB)

    Sadly I can't share my dataset, but it's about 1700 images (wihch make up for a total of 14GB)

    Here I share the configuration printed at the beggining of the process:

    [INFO] DTM is turned on, automatically turning on point cloud classification [INFO] Initializing ODM 2.8.1 - Mon Mar 21 15:28:30 2022 [INFO] ============== [INFO] auto_boundary: False [INFO] boundary: {} [INFO] build_overviews: False [INFO] camera_lens: auto [INFO] cameras: {} [INFO] cog: False [INFO] copy_to: None [INFO] crop: 3 [INFO] debug: False [INFO] dem_decimation: 1 [INFO] dem_euclidean_map: False [INFO] dem_gapfill_steps: 3 [INFO] dem_resolution: 2.0 [INFO] depthmap_resolution: 640 [INFO] dsm: True [INFO] dtm: True [INFO] end_with: odm_postprocess [INFO] fast_orthophoto: False [INFO] feature_quality: high [INFO] feature_type: sift [INFO] force_gps: False [INFO] gcp: None [INFO] geo: None [INFO] gps_accuracy: 10 [INFO] ignore_gsd: False [INFO] matcher_neighbors: 20 [INFO] matcher_type: flann [INFO] max_concurrency: 24 [INFO] merge: all [INFO] mesh_octree_depth: 11 [INFO] mesh_size: 200000 [INFO] min_num_features: 15000 [INFO] name: datasets [INFO] optimize_disk_space: False [INFO] orthophoto_compression: DEFLATE [INFO] orthophoto_cutline: False [INFO] orthophoto_kmz: False [INFO] orthophoto_no_tiled: False [INFO] orthophoto_png: False [INFO] orthophoto_resolution: 5 [INFO] pc_classify: True [INFO] pc_csv: False [INFO] pc_ept: False [INFO] pc_filter: 2.5 [INFO] pc_geometric: False [INFO] pc_las: False [INFO] pc_quality: medium [INFO] pc_rectify: False [INFO] pc_sample: 0 [INFO] pc_tile: False [INFO] primary_band: auto [INFO] project_path: /workspace [INFO] radiometric_calibration: none [INFO] rerun: None [INFO] rerun_all: True [INFO] rerun_from: None [INFO] resize_to: 2048 [INFO] sfm_algorithm: incremental [INFO] skip_3dmodel: False [INFO] skip_band_alignment: False [INFO] skip_orthophoto: False [INFO] skip_report: False [INFO] sm_cluster: None [INFO] smrf_scalar: 1.25 [INFO] smrf_slope: 0.15 [INFO] smrf_threshold: 0.5 [INFO] smrf_window: 18.0 [INFO] split: 999999 [INFO] split_image_groups: None [INFO] split_overlap: 150 [INFO] texturing_data_term: gmi [INFO] texturing_keep_unseen_faces: False [INFO] texturing_outlier_removal_type: gauss_clamping [INFO] texturing_skip_global_seam_leveling: False [INFO] texturing_skip_local_seam_leveling: False [INFO] texturing_tone_mapping: none [INFO] tiles: False [INFO] time: False [INFO] use_3dmesh: False [INFO] use_exif: False [INFO] use_fixed_camera_params: False [INFO] use_hybrid_bundle_adjustment: False [INFO] verbose: False [INFO] ============== [INFO] Rerun all -- Removing old data [INFO] Running dataset stage [INFO] Loading dataset from: /workspace/datasets/images [INFO] Loading 1734 images [INFO] Wrote images database: /workspace/datasets/images.json [INFO] Found 1734 usable images [INFO] Parsing SRS header: WGS84 UTM 35N [INFO] Finished dataset stage [INFO] Running split stage [INFO] Normal dataset, will process all at once. [INFO] Finished split stage [INFO] Running merge stage [INFO] Normal dataset, nothing to merge. [INFO] Finished merge stage [INFO] Running opensfm stage [INFO] Maximum photo dimensions: 4056px [INFO] Photo dimensions for feature extraction: 2028px [INFO] nvidia-smi detected [INFO] Using GPU for extracting SIFT features [INFO] Altitude data detected, enabling it for GPS alignment [INFO] ['use_exif_size: no', 'flann_algorithm: KDTREE', 'feature_process_size: 2028', 'feature_min_frames: 15000', 'processes: 24', 'matching_gps_neighbors: 20', 'matching_gps_distance: 0', 'matching_graph_rounds: 50', 'optimize_camera_parameters: yes', 'reconstruction_algorithm: incremental', 'undistorted_image_format: tif', 'bundle_outlier_filtering_type: AUTO', 'sift_peak_threshold: 0.066', 'align_orientation_prior: vertical', 'triangulation_type: ROBUST', 'retriangulation_ratio: 2', 'bundle_compensate_gps_bias: yes', 'matcher_type: FLANN', 'feature_type: SIFT_GPU', 'use_altitude_tag: yes', 'align_method: auto', 'local_bundle_radius: 0'] [INFO] Wrote reference_lla.json [INFO] running "/code/SuperBuild/install/bin/opensfm/bin/opensfm" detect_features "/workspace/datasets/opensfm" 2022-03-21 15:28:36,311 INFO: Capping memory usage to ~ 25173.736328125 MB 2022-03-21 15:28:36,311 INFO: Expecting to process 200 images.

    help wanted possible bug 
    opened by madisi98 2
  • SNAP: ModuleNotFoundError: No module named 'encodings'

    SNAP: ModuleNotFoundError: No module named 'encodings'

    This error was ignored in http://community.opendronemap.org/c/opendronemap so I open an issue here

    $ sudo /snap/bin/opendronemap 
    Python path configuration:
      PYTHONHOME = (not set)
      PYTHONPATH = (not set)
      program name = 'python3'
      isolated = 0
      environment = 1
      user site = 1
      import site = 1
      sys._base_executable = '/usr/bin/python3'
      sys.base_prefix = '/usr'
      sys.base_exec_prefix = '/usr'
      sys.executable = '/usr/bin/python3'
      sys.prefix = '/usr'
      sys.exec_prefix = '/usr'
      sys.path = [
        '/usr/lib/python38.zip',
        '/usr/lib/python3.8',
        '/usr/lib/python3.8/lib-dynload',
      ]
    Fatal Python error: init_fs_encoding: failed to get the Python codec of the filesystem encoding
    Python runtime state: core initialized
    ModuleNotFoundError: No module named 'encodings'
    
    Current thread 0x00007f66cf4d6740 (most recent call first):
    <no Python frame>
    
    $ uname -a; cat /etc/debian_version 
    Linux can 5.10.0-12-amd64 #1 SMP Debian 5.10.103-1 (2022-03-07) x86_64 GNU/Linux
    11.2
    
    help wanted possible bug 
    opened by sysdef 1
Releases(v2.8.6)
Owner
OpenDroneMap
Creating the most sustainable drone mapping software with the friendliest community on earth.
OpenDroneMap
A command-line tool for converting heightmaps in GeoTIFF format into tiled optimized meshes.

TIN Terrain TIN Terrain is a command-line tool for converting heightmaps presented in GeoTIFF format into tiled optimized meshes (Triangulated Irregul

HERE Technologies 465 Jun 30, 2022
OpenOrienteering Mapper is a software for creating maps for the orienteering sport.

OpenOrienteering Mapper OpenOrienteering Mapper is an orienteering mapmaking program and provides a free and open source alternative to existing comme

null 320 Jun 21, 2022
Terrain Analysis Using Digital Elevation Models (TauDEM) software for hydrologic terrain analysis and channel network extraction.

TauDEM (Terrain Analysis Using Digital Elevation Models) is a suite of Digital Elevation Model (DEM) tools for the extraction and analysis of hydrolog

David Tarboton 184 Jun 23, 2022
Entwine - point cloud organization for massive datasets

Build Status Entwine is a data organization library for massive point clouds, designed to conquer datasets of hundreds of billions of points as well a

Connor Manning 325 Jun 22, 2022
C++ implementation of R*-tree, an MVR-tree and a TPR-tree with C API

libspatialindex Author: Marios Hadjieleftheriou Contact: [email protected] Revision: 1.9.3 Date: 10/23/2019 See http://libspatialindex.org for full doc

null 614 Jun 20, 2022
2D and 3D map renderer using OpenGL ES

Tangram ES Tangram ES is a C++ library for rendering 2D and 3D maps from vector data using OpenGL ES. It is a counterpart to Tangram. This repository

Tangram 728 Jun 20, 2022
Alternative LAZ implementation for C++ and JavaScript

What is this? Alternative LAZ implementation. It supports compilation and usage in JavaScript, usage in database contexts such as pgpointcloud and Ora

Howard Butler 51 Jun 24, 2022
Computational geometry and spatial indexing on the sphere

S2 Geometry Library Overview This is a package for manipulating geometric shapes. Unlike many geometry libraries, S2 is primarily designed to work wit

Google 1.8k Jun 26, 2022
A C++17 image representation, processing and I/O library.

Selene Selene is a C++17 image representation, processing, and I/O library, focusing on ease of use and a clean, modern, type-safe API. Overview: Brie

Michael Hofmann 274 Jun 12, 2022
A fast algorithm for finding the pole of inaccessibility of a polygon (in JavaScript and C++)

A fast algorithm for finding polygon pole of inaccessibility, the most distant internal point from the polygon outline (not to be confused with centroid), implemented as a JavaScript library. Useful for optimal placement of a text label on a polygon.

Mapbox 1.2k Jun 27, 2022
A lean, efficient, accurate geohash encoder and decoder library implemented in C

Geohash encoder/decoder in C A lean, efficient, accurate geohash encoder and decoder library implemented in C. It does not depend on the C standard li

Christopher Wellons 18 May 30, 2022
A library of distance and occlusion generation routines

Distance/Occlusion Library + Tool From left to right: original, signed distance with zero at 0.5, red/green SDF, delta vectors to closest boundary poi

Andrew Willmott 98 Jan 3, 2022
Organic Maps is a better fork of MAPS.ME, an Android & iOS offline maps app for travelers, tourists, hikers, and cyclists based on top of crowd-sourced OpenStreetMap data and curated with love by MAPS.ME founders.

?? Organic Maps is a better fork of MAPS.ME, an Android & iOS offline maps app for travelers, tourists, hikers, and cyclists based on top of crowd-sourced OpenStreetMap data and curated with love by MAPS.ME founders. No ads, no tracking, no data collection, no crapware.

Organic Maps 3.5k Jun 25, 2022
LIDAR(Livox Horizon) point cloud preprocessing, including point cloud filtering and point cloud feature extraction (edge points and plane points)

LIDAR(Livox Horizon) point cloud preprocessing, including point cloud filtering and point cloud feature extraction (edge points and plane points)

hongyu wang 10 Apr 12, 2022
🗺️ OMAPS.APP — Offline OpenStreetMap maps for iOS and Android. A community-driven fork of MAPS.ME.

OMaps is an open source cross-platform offline maps application, built on top of crowd-sourced OpenStreetMap data. It was publicly released for iOS and Android.

OMaps 3.5k Jun 26, 2022
Draco is a library for compressing and decompressing 3D geometric meshes and point clouds.

Draco is a library for compressing and decompressing 3D geometric meshes and point clouds. It is intended to improve the storage and transmission of 3D graphics.

Google 5k Jun 30, 2022
Polyscope is a C++/Python viewer and user interface for 3D data such as meshes and point clouds

Polyscope is a C++/Python viewer and user interface for 3D data such as meshes and point clouds. It allows you to register your data and quickly generate informative and beautiful visualizations, either programmatically or via a dynamic GUI.

Nicholas Sharp 1.1k Jun 29, 2022
FG-Net: Fast Large-Scale LiDAR Point Clouds Understanding Network Leveraging Correlated Feature Mining and Geometric-Aware Modelling

FG-Net: Fast Large-Scale LiDAR Point Clouds Understanding Network Leveraging Correlated Feature Mining and Geometric-Aware Modelling Comparisons of Running Time of Our Method with SOTA methods RandLA and KPConv:

Kangcheng LIU 68 Jun 23, 2022
copc-lib provides an easy-to-use interface for reading and creating Cloud Optimized Point Clouds

copc-lib copc-lib is a library which provides an easy-to-use reader and writer interface for COPC point clouds. This project provides a complete inter

Rock Robotic 18 Jun 15, 2022
Lite.AI.ToolKit 🚀🚀🌟: A lite C++ toolkit of awesome AI models such as RobustVideoMatting🔥, YOLOX🔥, YOLOP🔥 etc.

Lite.AI.ToolKit ?? ?? ?? : A lite C++ toolkit of awesome AI models which contains 70+ models now. It's a collection of personal interests. Such as RVM, YOLOX, YOLOP, YOLOR, YoloV5, DeepLabV3, ArcFace, etc.

DefTruth 1.7k Jun 27, 2022
Direct LiDAR Odometry: Fast Localization with Dense Point Clouds

Direct LiDAR Odometry: Fast Localization with Dense Point Clouds DLO is a lightweight and computationally-efficient frontend LiDAR odometry solution w

VECTR at UCLA 244 Jun 21, 2022
NeeDrop: Self-supervised Shape Representation from Sparse Point Clouds using Needle Dropping

NeeDrop: Self-supervised Shape Representation from Sparse Point Clouds using Needle Dropping by: Alexandre Boulch, Pierre-Alain Langlois, Gilles Puy a

valeo.ai 25 May 24, 2022
Lidar-with-velocity - Lidar with Velocity: Motion Distortion Correction of Point Clouds from Oscillating Scanning Lidars

Lidar with Velocity A robust camera and Lidar fusion based velocity estimator to undistort the pointcloud. This repository is a barebones implementati

ISEE Research Group 128 Jun 20, 2022
C++ Kite Connect API library / SDK

Kite Connect API C++ client Overview Dependencies Getting dependencies Linux Others & uWS v0.14 Building & installation Examples REST API Ticker Docum

Zerodha Technology 23 Jun 2, 2022
2021 NUSpace Balloon Satellite programme

Balloon-Satellite A team from the School of Science and Technology, Singapore. the team (left to right) : le yang, min gyu, jarrett, sean the satellit

le yang 1 Dec 8, 2021
A command-line tool to generate Linux manual pages from C source code.

mangen A command-line tool to generate Linux manual pages from C source code. Description mangen is, as said above, a program to generate Linux manual

null 2 Nov 15, 2021
pbr2gltf2 is a command line tool for converting PBR images to a glTF 2.0 material.

pbr2gltf2 is a command line tool for converting PBR images to a glTF 2.0 material. The tool is detecting depending on the filename, which PBR information is stored. It swizzles the images and does reassign the channels to a glTF 2.0 image. The tool stores the images plus a minimal, valid glTF 2.0 file containing the required material, textures and images.

UX3D GmbH 22 Jan 11, 2022
Icopack - A simple command line tool to create multi-frame ICO files from PNG source images

Optidash is a modern, AI-powered image optimization and processing API. We will drastically speed-up your websites and save you money on bandwidth and

Optidash AI 59 Jun 15, 2022
A simple program that converts images into normal maps

Advertisement If you are a Blender user, this addon I created might be interesting for you: NormalmapGenerator This is a program to generate normal-,

Simon Wendsche 336 Jun 25, 2022