Event-based Fast Visual-Inertial Odometry by Adaptive Slicing of Time Surface

Jun Lu,Hui Feng,Wenwei Liu,Bo Hu

DOI: https://doi.org/10.1109/iaeac59436.2024.10503865

2024-01-01

Abstract:Event cameras are asynchronous bio-inspired vision sensors, possessing advantages such as high output frequency and high dynamic range, enabling the output of valid information in challenging scenes. Current visual odometry (VO) pipelines based on event cameras primarily leverage the high dynamic range advantage of event cameras, but struggle to achieve a balance between system accuracy and output frequency. In this paper, we propose a fast visual-inertial odometry (VIO) pipeline with adaptive slicing of the Time Surface, leveraging the complementary advantages of event cameras and standard cameras to achieve accurate and fast motion estimator. To achieve this goal, we introduce the method of valid events adaptive slice and an asynchronous tightly-coupled strategy during the optimization phase between standard cameras and event cameras/IMU. Experimental results on public benchmark datasets demonstrate that our pipeline outperforms state-of-the-art event-based VIO pipelines in terms of both accuracy and speed.