realsenseai/librealsense

★ 8,761C++Audience · developerComplexity · 3/5LicenseSetup · moderate

Mindmap

mindmap
  root((LibRealSense))
    Camera streams
      Color video
      Depth maps
      IMU motion data
    Language wrappers
      C++ Python
      ROS C# Unity
    Applications
      Robotics drones
      3D scanning
      People tracking
    Tools included
      RealSense Viewer
      Depth Quality Tool
      Build from source

mindmap root((LibRealSense)) Camera streams Color video Depth maps IMU motion data Language wrappers C++ Python ROS C# Unity Applications Robotics drones 3D scanning People tracking Tools included RealSense Viewer Depth Quality Tool Build from source

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Read live depth and color frames from a RealSense camera in a Python script for robotics or computer vision projects.

USE CASE 2

Build a 3D scanning application that captures depth maps and converts them to point clouds or 3D meshes.

USE CASE 3

Track people or objects in a room using depth data for interactive installations or autonomous navigation.

USE CASE 4

Integrate RealSense depth streams into a ROS robotics project for obstacle detection and mapping.

Tech stack

C++PythonROSC#UnityCMake

Getting it running

Difficulty · moderate Time to first run · 30min

Requires a physical RealSense camera, precompiled SDK packages make setup straightforward but source builds need additional system libraries.

Apache 2.0 license, use freely for any purpose, including commercial, as long as you preserve copyright and license notices.

In plain English

LibRealSense is the official open-source SDK for RealSense depth cameras. These are physical cameras that use stereo vision to measure how far away objects are, producing both a regular color image and a depth map at the same time. The library gives your program access to those streams and provides calibration data about the camera. The SDK is written in C++ and works on Windows, Linux, macOS, Android, and Docker. Beyond the C++ API, it includes wrappers for Python, ROS, C#, Unity, and several other environments. This means you can pull depth and color frames from a connected camera using whichever language or framework your project already uses. The camera hardware covers a range of models in the RealSense D400 series. Some are designed for close-up work with sub-millimeter accuracy, others for long-range sensing up to several meters. Certain models include a built-in IMU (motion sensor), and newer ones support Power over Ethernet or ruggedized GMSL connectors for industrial and robotics environments. Common applications shown in the documentation include robotics, drone navigation, 3D scanning, people tracking, and facial authentication. Getting started is straightforward if you download the precompiled SDK from the releases page. After connecting a RealSense camera, you can open the included RealSense Viewer application to see live depth and color streams, adjust settings, and record footage. A Depth Quality Tool is also included for checking measurement accuracy. Developers who need to build from source or integrate the library into an existing project will find setup guides for Linux, Windows, macOS, and Jetson hardware in the repository documentation. Note: this repository was recently migrated from the IntelRealSense organization to the RealSenseAI organization on GitHub.

Copy-paste prompts

Prompt 1

Using the librealsense Python wrapper, write a script that opens a RealSense D435 camera, captures 30 aligned color and depth frames, and saves them as PNG files.

Prompt 2

How do I read the distance in meters to a specific pixel in a RealSense depth frame using the Python API?

Prompt 3

I want to display a live colorized depth stream from a RealSense camera in an OpenCV window. Show me the minimal Python code to do that.

Prompt 4

How do I convert a RealSense depth frame to a 3D point cloud and export it to a PLY file using the librealsense Python bindings?

Prompt 5

Set up a ROS node that subscribes to a RealSense D435i depth stream and publishes obstacle distances for a robot navigation stack.

Open on GitHub → Explain another repo

← realsenseai on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.