Installation and Getting Started

This section is excerpted and modified with permission from OpenDroneMap: The Missing Guide, by Piero Toffanin.

Until recently OpenDroneMap was the term used to refer to a single command line application (what is now known as the ODM project). With the recent development of a web interface, an API and other tools, OpenDroneMap has become an ecosystem of various applications to process, analyze and display aerial data. This ecosystem is made of several components:

ODM Logo
  • ODM is the processing engine, which can be used from the command line. It takes images as input and produces a variety of outputs, including point clouds, 3D models and orthophotos

NodeODM Logo
  • NodeODM is a light-weight API built on top of ODM. It allows users and applications to access the functions of ODM over a computer network

WebODM Logo
  • WebODM is a friendly user interface that includes a map viewer, a 3D viewer, user logins, a plugin system and many other features that are expected of modern drone mapping platforms

CloudODM Logo
  • CloudODM is a small command line client to communicate with ODM via the NodeODM API

PyODM Logo
  • PyODM is a Python SDK for creating tasks via the NodeODM API. We cover it in more detail in the “Automated Processing With Python” chapter

ClusterODM Logo
  • ClusterODM is a load balancer for connecting together multiple NodeODM instances

ODM, NodeODM and WebODM are available on all major platforms (Windows, macOS and Linux) via a program called docker, which is required to run the software. Docker offers a way to run “containers”. Containers are packaged copies of an entire system, its software and its dependencies. These containers run within a virtual environment. On Linux this virtual environment is available from the operating system and is very efficient. On macOS and Windows the containers run within a VM, so there’s a bit of a overhead. but it’s still very suitable for running the software. Once installed users do not have to worry much about docker, as it operates (almost) transparently.

Without docker it would not be possible to run ODM on Windows or macOS. On these platforms ODM cannot run natively. Future development efforts are being focused on leveraging the new Windows Subsystem for Linux (WSL) and the possibility to make a native port of all dependencies to macOS, which is going to make the installation much easier.

On Ubuntu Linux 18.04 it’s feasible to run all OpenDroneMap software natively. However, because there’s very little performance penalty for running docker on Linux and docker is straightforward to setup on this platform, we don’t recommend it. On Linux the advantages of containerization far outweigh a tiny performance penalty. With docker users also get easy one-step updates of the software, so that’s nice.

Hardware Recommendations

The bare minimum requirements for running the software are:

  • 64bit CPU manufactured on or after 2010

  • 20 GB of disk space

  • 4 GB RAM

No more than 100-200 images can be processed with the above specifications (the software will run out of memory). Recommended requirements are:

  • Latest Generation CPU

  • 100 GB of disk space

  • 16 GB RAM

The above will allow for a few hundred images to be processed without too many issues. A CPU with more cores will allow for faster processing, while a graphics card (GPU) currently has no impact on performance. For processing more images, add more disk space and RAM linearly to the number of images you need to process.


We recommend people use docker for running ODM, whether you are on Windows, macOS or Linux.


To run OpenDroneMap you need at least Windows 7. Previous versions of Windows are not supported.

Step 1. Check Virtualization Support

Docker requires a feature from your CPU called virtualization, which allows it to run virtual machines (VMs). Make sure you have it enabled! Sometimes this is disabled. To check, on Windows 8 or higher you can open the Task Manager (press CTRL+SHIFT+ESC) and switch to the Performance tab.

Image of checking virtualization in Windows 8 or higher

Virtualization should be enabled

On Windows 7 to see if you have virtualization enabled you can use the Microsoft® Hardware-Assisted Virtualization Detection Tool instead.

If virtualization is disabled, you’ll need to enable it. The procedure unfortunately is a bit different for each computer model, so the best way to do this is to look up on a search engine “how to enable vtx for <type your computer model here>”. Often times it’s a matter of restarting the computer, immediately pressing F2 or F12 during startup, navigating the boot menu and changing the settings to enable virtualization (often called “VT-X”).

Table of different bios keys

Common keys to press at computer startup to access the boot menu for various PC vendors

Step 2. Install Requirements

First, you’ll need to install:

For Python 3, make sure you check Add Python 3.x to PATH during the installation.

Screenshot of Python3 installation process

Don’t forget to add the Python executable to your PATH (so that you can run commands with it)

Then, only if you are on Windows 10 Home, Windows 8 (any version) or Windows 7 (any version), install:

If you are on Windows 10 Professional or a newer version, you should install instead:

Please do NOT install both docker programs. They are different and will create a mess if they are both installed.

After installing docker, launch it from the Desktop icon that is created from the installation (Docker Quickstart in the case of Docker Toolbox, Docker for Windows for Docker for Windows). This is important, do not skip this step. If there are errors, follow the prompts on screen to fix them.

Step 3. Check Memory and CPU Allocation

Docker on Windows works by running a VM in the background (think of a VM as a “computer emulator”). This VM has a certain amount of memory allocated and WebODM can only use as much memory as it’s allocated.

If you installed Docker Toolbox (see below if you installed Docker for Windows instead):

  1. Open the VirtualBox Manager application

  2. Right click the default VM and press Close (ACPI Shutdown) to stop the machine

  3. Right click the default VM and press Settings…

  4. Move the Base Memory slider from the System paneland allocate 60-70% of all available memory, optionally adding 50% of the available processors from the Processor tab also

Screenshot of VirtualBox Settings

VirtualBox default VM settings

Then press OK, right click the default VM and press Start.

If you installed Docker for Windows instead:

  1. Look in the system tray and right click the “white whale” icon.

  2. From the menu, press Settings…

  3. From the panel, click Advanced and use the sliders to allocate 60-70% of available memory and use half of all available CPUs.

  4. Press Apply.

Screenshot of Docker Icon

Step 1 Docker icon

Screenshot of Docker Settings

Step 3 & 4 Docker settings

Step 4. Download WebODM

Open the Git Gui program that comes installed with Git. From there:

  • When Git Gui opens, click ‘Clone Existing Repository’ option

  • In Source Location type:

  • In Target Directory click browse and navigate to a folder of your choosing (create one if necessary)

  • Press Clone

Screenshot of Git Gui

Git Gui

If the download succeeded, you should now see this window:

Screenshot of Git Gui after successful download

Git Gui after successful download (clone)

Go to the Repository menu, then click Create Desktop Icon. This will allow you to come back to this application easily in the future.

Step 4. Launch WebODM

From Git Gui, go to the Repository menu, then click Git Bash. From the command line terminal type:

$ ./ start&

Several components will download to your machine at this point, including WebODM, NodeODM and ODM. After the download you should be greeted by the following screen:

Screenshot of after successfully downloading WebODM

Console output after starting WebODM for the first time

  • If you are using Docker for Windows, open a web browser to http://localhost:8000

  • If you are using Docker Toolbox, find the IP address to connect to by typing:

$ docker-machine ip

You should get a result like the following:

Then connect to (replacing the IP address with the proper one).


Most modern (post 2010) Mac computers running macOS Sierra 10.12 or higher can run OpenDroneMap using docker, as long as hardware virtualization is supported (see below).

Step 1. Check Virtualization Support

Open a Terminal window and type:

$ sysctl kern.hv_support

You will get a response similar to the following:

kern.hv_support: 1

If the result is kern.hv_support: 1, then your Mac is supported! Continue with Step 2.

If the result is kern.hv_support: 0, unfortunately it means your Mac is too old to run OpenDroneMap. :(

Step 2. Install Requirements

There are only two programs to install:

  1. Docker:

  2. Git:

After installing docker you should find an icon that looks like a whale in the task bar.

Screenshot of Docker whale

Docker app running

You can verify that docker is running properly by opening the Terminal app and typing:

$ docker run hello-world

Which should return

Hello from Docker!

To verify that git is installed, simply type:

$ git --version

Which should return something similar to the following:

git version 2.20.1 (Apple Git-117)

If you get a “bash: git: command not found”, try to restart your Terminal app and double-check for any errors during the install process.

Step 3. Check Memory and CPU Allocation

Docker on macOS works by running a VM in the background (think of it as a “computer emulator”). This VM has a certain amount of memory allocated and WebODM can only use as much memory as it’s allocated.

  1. Right click the whale icon from the task bar and click Preferences

  2. Select the Advanced tab

  3. Adjust the CPUs slider to use half of all available CPUs and the memory to use 60-70% of all available memory

  4. Press Apply & Restart

Screenshot of Docker advanced settings

Docker advanced settings

Step 4. Download and Launch WebODM

From a Terminal type:

$ git clone
$ cd WebODM
$ ./ start

Then open a web browser to http://localhost:8000.


OpenDroneMap can run on any Linux distribution that supports docker. According to docker’s documentation website the officially supported distributions are CentOS, Debian, Ubuntu and Fedora, with static binaries available for others. If you have to pick a distribution solely for running OpenDroneMap, Ubuntu is the recommended way to go.

Step 1. Install Requirements

There are four programs that need to be installed:

  1. Docker

  2. Git

  3. Python (2 or 3)

  4. Pip

We cannot possibly cover the installation process for every Linux distribution out there, so we’ll limit the instructions to those that are distributions officially supported by docker. In all cases it’s just a matter of opening a terminal prompt and typing a few commands.

Install on Ubuntu / Debian

Commands to type:

$ sudo apt update
$ curl -fsSL -o
$ sh
$ sudo apt install -y git python python-pip
Install on CentOS / RHEL

Commands to type:

$ curl -fsSL -o
$ sh
$ sudo yum -y install git python python-pip
Install on Fedora

Commands to type:

$ curl -fsSL -o
$ sh
$ sudo dnf install git python python-pip
Install on Arch

Commands to type:

$ sudo pacman -Sy docker git python python-pip

Step 2. Check Additional Requirements

In addition to the three programs above, the dockercompose script is also needed. Sometimes it’s already installed with docker, but sometimes it isn’t. To verify if it’s installed try to type:

$ docker-compose --version

You should see somethings similar to the following:

docker-compose version 1.22.0, build f46880f

If instead you get something similar to the following:

docker-compose: command not found

you can install it by using pip:

$ sudo pip install docker-compose

Step 3. Download and Launch WebODM

From a terminal type:

$ git clone
$ cd WebODM
$ ./ start

Then open a web browser to http://localhost:8000.

Basic Commands and Troubleshooting

The cool thing about using docker is that 99% of the tasks you’ll ever need to perform while using WebODM can be done via the ./ script. You have already encountered one of them:

$ ./ start

which takes care of starting WebODM and setting up a default processing node (node-odm-1). If you want to stop WebODM, you can already guess what the command is:

$ ./ stop

There are several other commands you can use, along with different flags. Flags are parameters passed to the ./ command and are typically prefixed with “–”. The port flag for example instructs WebODM to use a different network port:

$ ./ start --port 80

Other useful commands are listed below:

# Restart WebODM (useful if things get stuck)
$ ./ restart

# Reset the admin user's password if you forget it
$ ./ resetadminpassword newpass

# Update everything to the latest version
$ ./ update

# Store processing results in the specified folder instead of the default location (inside docker)
$ ./ restart --media-dir /path/to/webodm_results

# See all options
$ ./ --help

The community forum is a great place to ask for help if you get stuck during any of the installation steps and for general questions on using the ./ script.

Hello, WebODM!

After running ./ start and opening WebODM in the browser, you will be greeted with a welcome message and will be asked to create the first user. Take some time to familiarize yourself with the web interface and explore its various menus.

Screenshot of WebODM Dashboard

WebODM Dashboard

Notice that under the Processing Nodes menu there’s a “node-odm-1” node already configured for you to use. This is a NodeODM node and has been created automatically by WebODM. This node is running on the same machine as WebODM.

If you’ve made it this far, congratulations! Now it’s time to start processing some data.

Image of celebratory dance

Running on more than one machine

Optionally: If you have another computer, you can repeat the installation process (install docker, git, python, etc.) and launch a new NodeODM node by typing from a Terminal/Git Bash window:

docker run --rm -it -p 3000:3000 opendronemap/nodeodm -q 1 --token secret

The above command asks docker to launch a new container using the opendronemap/nodeodm image from Docker Hub (the latest version of NodeODM), using port 3000, setting a maximum number of concurrent tasks to 1 and to protect the node from unauthorized access using the password “secret”.

From WebODM you can then press the Add New button under Processing Nodes. For the hostname/IP field type the IP of the second computer. For the port field type “3000”. For the token field type “secret”. You can also add an optional label for your node, such as “second computer”. Then press Save.

If everything went well, you should now have two processing nodes! You will be able to process multiple tasks in parallel using two different machines.

Help edit these docs!