search

dronebird / WebODM

📷 ✈ User-friendly, extendable application and API for processing aerial imagery.
https://www.opendronemap.org/webodm/
Mozilla Public License 2.0
0 stars 0 forks source link

readme

WebODM

Build Status GitHub version

A user-friendly, extendable application and API for drone image processing. Generate georeferenced maps, point clouds, elevation models and textured 3D models from aerial images. It supports multiple engines for processing, currently ODM and MicMac (experimental).

image

image

image

Getting Started

Docker Toolbox users need to find the IP of their docker machine by running this command from the Docker Quickstart Terminal:

docker-machine ip
192.168.1.100 (your output will be different)

The address to connect to would then be: http://192.168.1.100:8000.

To stop WebODM press CTRL+C or run:

./webodm.sh stop

To update WebODM to the latest version use:

./webodm.sh update

We recommend that you read the Docker Documentation to familiarize with the application lifecycle, setup and teardown, or for more advanced uses. Look at the contents of the webodm.sh script to understand what commands are used to launch WebODM.

Windows and macOS users can purchase an automated installer, which makes the installation process easier.

You can also run WebODM from a Live USB/DVD. See LiveODM.

Manage Processing Nodes

WebODM can be linked to one or more processing nodes that speak the NodeODM API, such as NodeODM, NodeMICMAC or ClusterODM. The default configuration includes a "node-odm-1" processing node which runs on the same machine as WebODM, just to help you get started. As you become more familiar with WebODM, you might want to install processing nodes on separate machines.

Adding more processing nodes will allow you to run multiple jobs in parallel.

You can also setup a ClusterODM node to run a single task across multiple machines with distributed split-merge and process dozen of thousands of images more quickly, with less memory.

If you don't need the default "node-odm-1" node, simply pass --default-nodes 0 flag when starting WebODM:

./webodm.sh restart --default-nodes 0.

Then from the web interface simply manually remove the "node-odm-1" node.

Enable MicMac

WebODM can use MicMac as a processing engine via NodeMICMAC. To add MicMac, simply run:

./webodm.sh restart --with-micmac

This will create a "node-micmac-1" processing node on the same machine running WebODM. Please note that NodeMICMAC is in active development and is currently experimental. If you find issues, please report them on the NodeMICMAC repository.

Enable SSL

WebODM has the ability to automatically request and install a SSL certificate via Let’s Encrypt, or you can manually specify your own key/certificate pair.

./webodm.sh restart --ssl --hostname webodm.myorg.com

That's it! The certificate will automatically renew when needed.

If you want to specify your own key/certificate pair, simply pass the --ssl-key and --ssl-cert option to ./webodm.sh. See ./webodm.sh --help for more information.

Where Are My Files Stored?

When using Docker, all processing results are stored in a docker volume and are not available on the host filesystem. If you want to store your files on the host filesystem instead of a docker volume, you need to pass a path via the --media-dir option:

./webodm.sh restart --media-dir /home/user/webodm_data

Note that existing task results will not be available after the change. Refer to the Migrate Data Volumes section of the Docker documentation for information on migrating existing task results.

Common Troubleshooting

Symptoms Possible Solutions
While starting WebODM you get: from six.moves import _thread as thread ImportError: cannot import name _thread Try running: sudo pip install --ignore-installed six
While starting WebODM you get: 'WaitNamedPipe','The system cannot find the file specified.' 1. Make sure you have enabled VT-x virtualization in the BIOS.
2. Try to downgrade your version of Python to 2.7
While Accessing the WebODM interface you get: OperationalError at / could not translate host name “db” to address: Name or service not known or ProgrammingError at / relation “auth_user” does not exist Try restarting your computer, then type: ./webodm.sh restart
Task output or console shows one of the following:
  • MemoryError
  • Killed
 Make sure that your Docker environment has enough RAM allocated: MacOS Instructions, Windows Instructions
After an update, you get: django.contrib.auth.models.DoesNotExist: Permission matching query does not exist. Try to remove your WebODM folder and start from a fresh git clone
Task fails with Process exited with code null, no task console output - OR - console output shows Illegal Instruction - OR - console output shows Child returned 132 If the computer running NodeODM is using an old or 32bit CPU, you need to compile OpenDroneMap from sources and setup NodeODM natively. You cannot use docker. Docker images work with CPUs with 64-bit extensions, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2 and POPCNT instruction set support or higher.
On Windows, docker-compose fails with Failed to execute the script docker-compose Make sure you have enabled VT-x virtualization in the BIOS
Cannot access WebODM using Microsoft Edge on Windows 10 Try to tweak your internet properties according to these instructions
Getting a No space left on device error, but hard drive has enough space left Docker on Windows by default will allocate only 20GB of space to the default docker-machine. You need to increase that amount. See this link and this link
Cannot start WebODM via ./webodm.sh start, error messages are different at each retry You could be running out of memory. Make sure you have enough RAM available. 2GB should be the recommended minimum, unless you know what you are doing
While running WebODM with Docker Toolbox (VirtualBox) you cannot access WebODM from another computer in the same network. As Administrator, run cmd.exe and then type "C:\Program Files\Oracle\VirtualBox\VBoxManage.exe" controlvm "default" natpf1 "rule-name,tcp,,8000,,8000"

Have you had other issues? Please report them so that we can include them in this document.

Backup and Restore

If you want to move WebODM to another system, you just need to transfer the docker volumes (unless you are storing your files on the file system).

On the old system:

mkdir -v backup
docker run --rm --volume webodm_dbdata:/temp --volume `pwd`/backup:/backup ubuntu tar cvf /backup/dbdata.tar /temp
docker run --rm --volume webodm_appmedia:/temp --volume `pwd`/backup:/backup ubuntu tar cvf /backup/appmedia.tar /temp

Your backup files will be stored in the newly created backup directory. Transfer the backup directory to the new system, then on the new system:

ls backup # --> appmedia.tar  dbdata.tar
./webodm.sh down # Make sure WebODM is down
docker run --rm --volume webodm_dbdata:/temp --volume `pwd`/backup:/backup ubuntu bash -c "rm -fr /temp/* && tar xvf /backup/dbdata.tar"
docker run --rm --volume webodm_appmedia:/temp --volume `pwd`/backup:/backup ubuntu bash -c "rm -fr /temp/* && tar xvf /backup/appmedia.tar"
./webodm.sh start

Reset Password

If you forgot the password you picked the first time you logged into WebODM, to reset it just type:

./webodm.sh start && ./webodm.sh resetadminpassword newpass

The password will be reset to newpass. The command will also tell you what username you chose.

Manage Plugins

Plugins can be enabled and disabled from the user interface. Simply go to Administration -- Plugins.

Update

If you use docker, updating is as simple as running:

./webodm.sh update

If you are running WebODM natively, these commands should do it:

cd /webodm
sudo su odm # Only in case you are running WebODM with a different user
git pull origin master
source python3-venv/bin/activate # If you are running a virtualenv
npm install
pip install -r requirements.txt
webpack --mode production
python manage.py collectstatic --noinput
python manage.py migrate

Recommended Machine Specs

To run a standalone installation of WebODM (the user interface), including the processing component (NodeODM), we recommend at a minimum:

Don't expect to process more than a few hundred images with these specifications. To process larger datasets, add more RAM linearly to the number of images you want to process. A CPU with more cores will speed up processing, but can increase memory usage. GPU acceleration is still a work in progress, so currently a good video card does not improve performance.

WebODM runs best on Linux, but works well on Windows and Mac too. If you are technically inclined, you can get WebODM to run natively on all three platforms and there's a native installer for Ubuntu 16.04 also available.

NodeODM and ODM cannot run natively on Mac and Windows and this is the reason we mostly recommend people to use docker.

WebODM by itself is just a user interface (see below) and does not require many resources. WebODM can be loaded on a machine with just 1 or 2 GB of RAM and work fine without NodeODM. You can then use a processing service such as the lightning network or run NodeODM on a separate, more powerful machine.

Customizing and Extending

Small customizations such as changing the application colors, name, logo, or addying custom CSS/HTML/Javascript can be performed directly from the Customize -- Brand/Theme panels within WebODM. No need to fork or change the code.

More advanced customizations can be achieved by writing plugins. This is the preferred way to add new functionality to WebODM since it requires less effort than maintaining a separate fork. The plugin system features server-side signals that can be used to be notified of various events, a ES6/React build system, a dynamic client-side API for adding elements to the UI, a built-in data store, an async task runner, a GRASS engine, hooks to add menu items and functions to rapidly inject CSS, Javascript and Django views.

The plugin system is still in beta. The best source of documentation currently is to look at existing code. If a particular hook / entrypoint for your plugin does not yet exist, request it. We are adding hooks and entrypoints as we go.

To create a plugin simply copy the plugins/test plugin into a new directory (for example, plugins/myplugin), then modify manifest.json, plugin.py and issue a ./webodm.sh restart.

API Docs

See the API documentation page.

ODM, NodeODM, WebODM... what?

The OpenDroneMap project is composed of several components.

webodm

In general, follow these guidelines to find out what you should use:

I am a... Best choice
End user, I'm not really comfortable with the command line WebODM
End user, I like shell commands, I need to process images for myself. I use other software to display processing results ODM
End user, I can work with the command line, but I'd rather not. I use other software to display processing results NodeODM
End user, I need a drone mapping application for my organization that everyone can use. WebODM
Developer, I'm looking to build an app that displays map results and takes care of things like permissions WebODM
Developer, I'm looking to build an app that will stay behind a firewall and just needs raw results NodeODM

Roadmap

We follow a bottom-up approach to decide what new features are added to WebODM. User feedback guides us in the decision making process and we collect such feedback via improvement requests.

Don't see a feature that you want? Open a feature request or help us build it.

Sometimes we also prioritize work that has received financial backing. If your organization is in the position to financially support the development of a particular feature, get in touch and we'll make it happen.

Getting Help

We have several channels of communication for people to ask questions and to get involved with the community:

We also have a Gitter Chat, but the preferred way to communicate is via the OpenDroneMap Community Forum.

Support the Project

There are many ways to contribute back to the project:

Become a Contributor

The easiest way to get started is to take a look at our list of outstanding issues and pick one. You can also fix/improve something entirely new based on your experience with WebODM. All ideas are considered and people of all skill levels are welcome to contribute.

You don't necessarily need to be a developer to become a contributor. We can use your help to write better documentation and improve the user interface texts and visuals.

If you know how to code, we primarily use Python (Django), Javascript (React), HTML and SCSS. See the Development Quickstart and Contributing documents for more information.

To make a contribution, you will need to open a pull request (here's how). To make changes to WebODM, make a clone of the repository and run ./webodm.sh start --dev.

If you have questions visit us on the forum and we'll be happy to help you out with your first contribution.

Architecture Overview

WebODM is built with scalability and performance in mind. While the default setup places all databases and applications on the same machine, users can separate its components for increased performance (ex. place a Celery worker on a separate machine for running background tasks).

Architecture

A few things to note:

Run the docker version as a Linux Service

If you wish to run the docker version with auto start/monitoring/stop, etc, as a systemd style Linux Service, a systemd unit file is included in the service folder of the repo.

This should work on any Linux OS capable of running WebODM, and using a SystemD based service daemon (such as Ubuntu 16.04 server for example).

This has only been tested on Ubuntu 16.04 server.

The following pre-requisites are required:

If all pre-requisites have been met, and repository is checked out to /opt/WebODM folder, then you can use the following steps to enable and manage the service:

First, to install the service, and enable the services to run at startup from now on:

sudo systemctl enable /webodm/service/webodm-gunicorn.service
sudo systemctl enable /webodm/service/webodm-nginx.service

To manually start/stop the service:

sudo systemctl stop webodm-gunicorn
sudo systemctl start webodm-gunicorn

To manually check service status:

sudo systemctl status webodm-gunicorn

Run it natively

WebODM can run natively on Windows, MacOS and Linux. We don't recommend to run WebODM natively (using docker is easier), but it's possible.

Ubuntu 16.04 LTS users can refer to this community script to install WebODM natively on a new machine.

To run WebODM, you will need to install:

On Linux, make sure you have:

apt-get install binutils libproj-dev gdal-bin nginx

On Windows use the OSGeo4W installer to install GDAL. MacOS users can use:

brew install postgres postgis

Then these steps should be sufficient to get you up and running:

git clone --depth 1 https://github.com/OpenDroneMap/WebODM

Create a WebODM/webodm/local_settings.py file containing your database settings:

DATABASES = {
    'default': {
        'ENGINE': 'django.contrib.gis.db.backends.postgis',
        'NAME': 'webodm_dev',
        'USER': 'postgres',
        'PASSWORD': 'postgres',
        'HOST': 'localhost',
        'PORT': '5432',
    }
}

From psql or pgadmin, connect to PostgreSQL, create a new database (name it webodm_dev), connect to it and set the postgis.enable_outdb_rasters and postgis.gdal_enabled_drivers settings:

ALTER SYSTEM SET postgis.enable_outdb_rasters TO True;
ALTER SYSTEM SET postgis.gdal_enabled_drivers TO 'GTiff';

Start the redis broker:

redis-server

Then:

pip install -r requirements.txt
sudo npm install -g webpack
sudo npm install -g webpack-cli
npm install
webpack --mode production
python manage.py collectstatic --noinput
chmod +x start.sh && ./start.sh --no-gunicorn

Finally, start at least one celery worker:

./worker.sh start

The start.sh script will use Django's built-in server if you pass the --no-gunicorn parameter. This is good for testing, but bad for production.

In production, if you have nginx installed, modify the configuration file in nginx/nginx.conf to match your system's configuration and just run start.sh without parameters.

Windows users should refer to this guide to install Apache + mod_wsgi and run gunicorn:

gunicorn webodm.wsgi --bind 0.0.0.0:8000 --preload

If you are getting a rt_raster_gdal_warp: Could not create GDAL transformation object for output dataset creation, make sure that your PostGIS installation has PROJ support:

SELECT PostGIS_Full_Version();

You may also need to set the environment variable PROJSO to the .so or .dll projection library your PostGIS is using. This just needs to have the name of the file. So for example on Windows, you would in Control Panel -> System -> Environment Variables add a system variable called PROJSO and set it to libproj.dll (if you are using proj 4.6.1). You'll have to restart your PostgreSQL service/daemon after this change. http://postgis.net/docs/manual-2.0/RT_ST_Transform.html

If you are using Windows and are unable to go past the pip install -r requirements.txt command because of an error regarding zlib and Pillow, manually edit the requirements.txt file, remove the Pillow requirement and run:

easy_install pillow
pip install -r requirements.txt

On Windows make sure that all of your PATH environment variables are set properly. These commands:

python --version
pip --version
npm --version
gdalinfo --version
redis-server --version

Should all work without errors.

Run it on the cloud (Google Compute, Amazon AWS)

12 steps, to have WebODM running on a cloud instance.

These steps are for Google Cloud, but can also be used for Amazon AWS, and other cloud platforms with small modifications:

  1. Launch a Google Cloud instance of Ubuntu 18.0 LTS.
  2. Open the SSH terminal - Google offers SSH via the website.
  3. Run sudo apt-get update
  4. Run sudo apt-get upgrade
  5. Run sudo apt-get install docker-compose
  6. Run sudo apt-get install python-pip
  7. Run git clone https://github.com/OpenDroneMap/WebODM --config core.autocrlf=input --depth 1
  8. cd WebODM (Linux is case sensitive)
  9. sudo ./webodm.sh start
  10. You now can access webodm via the public IP address for your google instance. Remember the default port of 8000.
  11. Check that your instance's firewall is allowing inbound TCP connections on port 8000! If you forget this step you will not be able to connect to WebODM.
  12. Open http://GooglepublicIPaddressforyourinstance:8000

To setup the firewall on Google Cloud, open the instance, on the middle of the instance settings page find NIC0. Open it, and then add the TCP Port 8000 for ingress, and egress on the Firewall.