3D LiDAR/IMU Calibration Based on Continuous-Time Trajectory Estimation in Structured Environments

Sensor calibration is a fundamental step for improving the performance in sensor fusion, aim of which to spatially and temporally register sensors with respect each other. This paper presents high-accuracy autocalibration method estimate extrinsic parameters between LiDAR an IMU. LiDAR/IMU challenging task since raw measurements are distorted, biased, noisy, asynchronous. Our approach adopts continuous-time trajectory estimation wherein IMU modeled by Gaussian process(GP) regression independent sampling timestamps. Accordingly, distorted delayed points sampled at discrete timestamps can be analytically on-manifold batch optimization. To efficiently accurately associate laser stable environmental objects, carried out known environments point map that segmented as structured planes managed specially designed octree map.We thoroughly investigated factors relevant accuracy evaluated proposed using both simulated real-world datasets. The results demonstrate robustness our sufficient most applications.