Robust and Efficient Vehicles Motion Estimation with Low-Cost Multi-Camera and Odometer-Gyroscope

Wenlong Ye,Renjie Zheng,Fangqiang Zhang,Zizhou Ouyang,Yong Liu

DOI: https://doi.org/10.1109/iros40897.2019.8968048

2019-01-01

Abstract:In this paper, we present a robust and efficient estimation approach with multi-camera, odometer and gyroscope. Robust initialization, tightly-coupled optimization estimator and multi-camera loop-closure detection are utilized in the proposed approach. In initialization, the measurements of odometer and gyroscope are used to compute scale, and then estimate the bias of sensors. In estimator, the pre-integration of odometer and gyroscope is derived and combined with the measurements of multi-camera to estimate the motion in a tightly-coupled optimization framework. In loop-closure detection, a connection between different cameras of the vehicle can be built, which significantly improve the success rate of loop-closure detection. The proposed algorithm is validated in multiple real-world datasets collected in different places, time, weather and illumination. Experimental results show that the proposed approach can estimate the motion of vehicles robustly and efficiently.