A ROS-based tool for manually calibrating extrinsic parameters between Livox LiDAR sensors and cameras using board corners.
Livox Camera-LiDAR Calibration is a ROS-based toolkit for manually calibrating the extrinsic parameters between Livox LiDAR sensors and cameras. It solves the problem of accurately aligning 3D point cloud data with 2D image data by using calibration board corners as reference targets, enabling precise sensor fusion for applications like autonomous vehicles and robotics.
Robotics engineers, autonomous vehicle researchers, and developers working with Livox LiDAR sensors who need to fuse camera and LiDAR data for perception systems.
It provides a specialized, manual calibration pipeline optimized for Livox LiDAR's non-repetitive scanning pattern, offering higher accuracy in corner detection from point clouds compared to generic methods, and includes verification tools for fusion quality assessment.
Calibrate the extrinsic parameters between Livox LiDAR and camera
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Leverages Livox LiDAR's non-repetitive scanning pattern to precisely locate calibration board corners in point clouds, as highlighted in the README for better fusion results.
Includes nodes for intrinsic calibration, data collection, extrinsic optimization (with two methods), and verification tools like point cloud projection and coloring, providing an end-to-end solution.
Offers UI tools for manual corner selection in images and point clouds, allowing users to ensure accuracy, though it requires careful data handling as noted in the steps.
Provides applications to project LiDAR point clouds onto images and colorize point clouds using camera data, enabling visual validation of calibration quality.
Requires installation of ROS, Livox SDK, PCL, Eigen, Ceres-solver, and optionally MATLAB, making initial configuration time-consuming and error-prone.
Demands manual corner coordinate extraction in both images and point clouds for each data point, increasing effort and potential for human error, as described in Steps 4.2 and 4.3.
Optimization results heavily depend on correctly setting initial extrinsic values; the README warns that bad initial values can lead to local optima or high costs.
Only verified for specific Livox LiDAR models and requires custom ROS message formats, restricting use in broader sensor setups or with other hardware.