A simplified high-precision intrinsic parameters calibration for multi-layer LiDAR using a 3D static target

Multi-layer mechanical LiDARs are core sensors for high-precision perception in autonomous driving, robotics, and 3D mapping. However, inherent systematic errors from assembly imperfections, TOF principle limitations, and component misalignments degrade key parameter accuracy (range, horizontal/vertical angular resolution) and point cloud consistency. To address this, a novel high-precision calibration method is proposed, integrating a 3D standard static target (cube) and high-precision laser rangefinders for intrinsic parameter correction. A systematic error mathematical model quantifies range and angular ( \(φ , θ \) ) deviations: the 3D target provides a geometric benchmark, while dual rangefinders offer ground truth to eliminate interference. Range calibration adopts a scaling coefficient method to align point clouds with the target’s ideal plane, and angular calibration uses edge point extraction and vector calculation for deviation correction. Experimental validation on Velodyne HDL-32E shows significant error reduction: average range deviation drops from 0.0077 m to 0.0028 m, horizontal angular error stabilizes at 0.0073 \(∘\) , and point cloud alignment RMSE reduces to 0.0065. Post-calibration errors follow a normal distribution ( \(1σ \) : 68.27%, \(3σ \) : 99.73% confidence intervals), confirming reliability. This low-cost, flexible method is adaptable to various multi-layer LiDARs, providing a practical solution for precision-dependent applications.