Precise Synchronization Between LiDAR and Multiple Cameras for Autonomous Driving: An Adaptive Approach

Abstract

LiDAR and camera are crucial perception sensors that provide complementary information for object detection in autonomous driving vehicles. However, the fusion of sensor data to achieve efficient object detection requires accurate calibration and precise time synchronization between the sensors. While calibration ensures the geometric relationship between the sensors, and time synchronization ensures that data from both sensors corresponds to the same moment in the real world. Even though various methods have been introduced to achieve accurate calibration between LiDAR and camera, time synchronization between sensors remains relatively unexplored. Poor synchronization between sensors caused by incorrect time stamping process significantly affects data fusion. Therefore, in this paper we present sensor time synchronization for LiDAR and camera data fusion approach, especially in autonomous driving vehicles. It also explores various techniques involved in establishing synchronization between LiDAR and camera. Subsequently, we propose a hardware level time synchronization system, including automatic hardware trigger signal delay estimation to precisely match LiDAR and camera trigger time. Furthermore, validation experiments were conducted to assess the accuracy of the proposed synchronization system in various driving scenarios accompanied by detailed experimental analysis.