"视觉惯性传感器时间同步标定方法及在SLAM技术中的应用"

Simultaneous localization and mapping (SLAM) based on visual-inertial information fusion is a crucial technology in the field of robot navigation. Synchronizing multimodal data is essential for data fusion algorithms, and obtaining accurate timestamps for different sensors is a key step in this process. While the timestamp for visual sensors can be accurately obtained using hardware methods, the timestamp for inertial sensors is usually replaced with output time, which can reduce the accuracy of visual-inertial fusion algorithms. To address this issue, a method for synchronizing the time of visual and inertial sensors is proposed. The method involves the design of a calibration device based on a planar pendulum, where visual and inertial sensors independently capture data during the motion of the pendulum in the plane. Both sensors are synchronized based on the same clock with added time markings. Next, a method based on least squares estimation is proposed to estimate the angular displacement and angular velocity functions of the pendulum bob. By comparing the phase difference between these two functions, the output delay and timestamp of the inertial sensor can be determined. Experimental results demonstrate that the proposed method has a mean square deviation of 0.018 ms for repeated calibrations, proving its effectiveness. This method provides a reliable and accurate way to synchronize the time of visual and inertial sensors, which is essential for improving the accuracy of SLAM algorithms based on visual-inertial fusion. By ensuring precise timing synchronization, robots can navigate more effectively and accurately map their surroundings, ultimately enhancing their overall performance and capabilities in various applications.