학술논문
SensorX2Vehicle: Online Sensors-to-Vehicle Rotation Calibration Methods in Road Scenarios
Document Type
Periodical
Source
IEEE Robotics and Automation Letters IEEE Robot. Autom. Lett. Robotics and Automation Letters, IEEE. 9(4):3775-3782 Apr, 2024
Subject
Language
ISSN
2377-3766
2377-3774
2377-3774
Abstract
Properly-calibrated sensors are the prerequisite for a dependable autonomous driving system. Besides the extrinsic calibration between the sensors, the extrinsic between the sensor and the vehicle is also important, especially the rotation. Most of the existing sensor-to-vehicle calibration approaches have requirements on facilities, manual work, road features or vehicle trajectory. In this work, we propose more general and flexible methods for four commonly used sensors: Camera, LiDAR, GNSS/INS, and millimeter-wave Radar, composing a toolbox named SensorX2car. In each method, the rotation between a sensor and the vehicle is calibrated individually in road scenarios. Experiments on large-scale real-world datasets demonstrate the practicality of our proposed methods. Meanwhile, the related codes have been open-sourced to benefit the community. To our knowledge, SensorX2car is the first open-source sensor to vehicle calibration toolbox.