Online Self-Calibration of Multiple 2D LiDARs Using Line Features With Fuzzy Adaptive Covariance

The calibration techniques for a multi-sensor data fusion need to deal with the high accuracy of each sensor position to the reference frame. The point cloud data generated by low-cost LiDAR sensors possess high uncertainty, giving poor affection to the calibration system. Furthermore, the displacements between the sensors could be adjusted by undesirable situations during operation time. This paper proposes an automatic calibration framework for multiple low-cost 2D lasers on the special Euclidean group SE(2). The key ideas of the proposed method are to perform the line feature-based factor graph optimization and fuzzy inference system to adapt the covariance of the factor graph. Here, to fuse the line feature information, we mimic the weight affection technique using the fuzzy logic rules and then add the adapted covariance to the graph. Next, we suggest an online self-calibration algorithm running in real-time to correct the variation of calibrated parameters. Lastly, the proposed calibration system is evaluated through a real-world dataset. The experiment results show that the proposed method can be widely applied for industrial applications.