CloudSoftwareHardware
Software Stack
DepthAI
OAK Apps
AI Models

OAK4

RVC4 based new generation OAK cameras

OAK

RVC2 based OAK cameras

Software

1

Getting Started

The easiest and quickest way to get started with your device.

Connecting to your OAK4 camera

Once you have completed the OAK4 Getting Started, you should be able to test out it's functionality.

OAK Viewer

OAK Viewer is a desktop GUI application that allows you to visualize the camera's streams and interact with the device. It is available for Windows, macOS, and Linux.
Click here to get started
2

Development

Ready-made applications

Try out our pre-made applications to dip your toes into our ecosystem. Here is an example of how to start the generic neural network app:Install oakctl
Linux/MacOS
Windows

Linux/MacOS

On 64bit system, run:
Command Line
1bash -c "$(curl -fsSL https://oakctl-releases.luxonis.com/oakctl-installer.sh)"
Deploy an example App
Command Line
1# Connect to the device
2oakctl connect
3# Clone the repo
4git clone --depth 1 --branch main https://github.com/luxonis/oak-examples.git
5# Change directory to the generic-example
6cd oak-examples/neural-networks/generic-example/
7# Run the app and follow instructions to open the web viewer
8oakctl app run .
Continue exploring other pre-made applications

Build your app from the ground up

Quickest way to get started with developing on your own is to use the template app. This app is a starting point for your own applications and includes all the necessary components to get you up and running.
Command Line
1# Install core packages
2pip install depthai --force-reinstall
3# Clone the template app
4git clone https://github.com/luxonis/oak-template.git
5# Change directory to the template app
6cd oak-template
7# Install requirements
8pip install -r requirements.txt
This will download the template / cookie-cutter app to your current directory.
Text
1template-app/
2├── main.py
3├── oakapp.toml
4├── requirements.txt
5├── media/ # Optional media files
6└── utils/ # Optional helper functions
Continue developing your template app

Ready-made applications

Try out our pre-made applications to dip your toes into our ecosystem. Here is an example of how to start the generic neural network app:Deploy an example App
Linux/MacOS
Windows

Linux/MacOS

Command Line
1# Clone the repo
2git clone --depth 1 --branch main https://github.com/luxonis/oak-examples.git
3# Change directory to the generic-example
4cd oak-examples/neural-networks/generic-example/
5# Create and enter virtual python environment
6python3 -m venv venv
7source venv/bin/activate
8# Install requirements into the environment
9pip install -r requirements.txt
10# Run the app and follow instructions to open the web viewer
11python3 main.py
Continue exploring other pre-made applications

Build your app from the ground up

Quickest way to get started with developing on your own is to use the template app. This app is a starting point for your own applications and includes all the necessary components to get you up and running.
Command Line
1# Install core packages
2pip install depthai --force-reinstall
3# Clone the template app
4git clone https://github.com/luxonis/oak-template.git
5# Change directory to the template app
6cd oak-template
7# Install requirements
8pip install -r requirements.txt
This will download the template / cookie-cutter app to your current directory.
Text
1template-app/
2├── main.py
3├── oakapp.toml
4├── requirements.txt
5├── media/ # Optional media files
6└── utils/ # Optional helper functions
Continue developing your template app
3

Ecosystem Overview

DepthAI

DepthAI is a powerful API that allows you to develop and deploy vision pipelines that run on the accelerated hardware blocks on our devices. It is available for Python and C++.
Click here to get started

DepthAI Nodes

Extend the functionality of your applications by using the pre-built nodes. Nodes can enable custom AI model output decoding, image processing, and more.
DepthAI Nodes Github

Model Zoo

Explore our model zoo to find the right model for your application. The model zoo contains a variety of models that can be used for different purposes like object detection, image classification, etc.
Model Zoo

Convert your custom AI model

Explore and convert AI models to run on our devices
Click here to get started
4

Key Concepts

For a deep dive into the software, it is important to understand a few key concepts about the device you are using.

OAK4 Topology

  • oak-agent: Device service that handles the communication and deployment of apps on the OAK4 device. It is controlled by the oakctl CLI tool. Good for mass deployment and remote management of OAK4 devices. Learn more.
  • DepthAI: The API library (Python/C++) that allows you to interact with the OAK4 sensors and hardware blocks. It is the main interface for the user to interact with the device. Learn more.
  • Luxonis OS: The operating system that runs on the OAK4 device. It is a custom Linux distro (Yocto based) that is designed to run on the OAK4 hardware. Learn more.
  • Hardware: The OAK4 device is based on the Qualcomm SoC and has a number of hardware blocks that can be used for different purposes like image processing, AI inference, etc.

Standalone vs Peripheral

  • Standalone: The device runs the entire application on the OAK4 device itself. This is useful when the user wants to deploy the device in the field, and doesn't want to rely on any additional computer. Apps are deployed via oakctl.
  • Peripheral: The device runs the app on the OAK4 device, but also wants to run some additional processing on the host computer. This is useful when the user wants to use the OAK4 device as a peripheral device and wants to use the host computer for additional processing. Peripheral mode apps are run using the DepthAI API.