ai-agent

ORB SLAM3

Name: ORB SLAM3
Author: UZ-SLAMLab

ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAM

8,892 stars3,138 forksC++Updated 7/24/2024100% free · open source

What it does

ORB_SLAM3 is an open-source library for visual, visual-inertial, and multi-map simultaneous localization and mapping (SLAM) that provides an accurate and efficient solution for augmented reality, robotics, and other applications

Install / run

git clone https://github.com/UZ-SLAMLab/ORB_SLAM3.git

When to use it

•When you need to track the location and orientation of a camera in real-time
•When you want to build an augmented reality application that requires precise localization
•When you need to perform SLAM on a dataset with multiple maps

Quick start

1Compile the library by running 'chmod +x build.sh' and then './build.sh' in the ORB_SLAM3 directory
2Create a configuration file for your dataset by modifying the 'config.ini' file in the 'Examples' directory
3Run the SLAM system using the command './Examples/mono_inertial_slam ./config.ini' for a monocular inertial setup
4Visualize the results using a tool like rviz orpcl_viewer, as described in the 'Examples' section of the README
5Tune the parameters in the configuration file to optimize the performance of the SLAM system for your specific use case

Ready-to-paste prompt

./Examples/mono_inertial_slam ./config.ini

Heads up: You need to have a compatible C++ compiler (e.g. GCC 9 or higher) and the necessary dependencies (e.g. OpenCV, Eigen) installed on your system before compiling ORB_SLAM3

Saves to your device