Scale-aware monocular visual-inertial pose estimation for aerial robots

Abstract

Monocular visual odometry algorithm has been widely used to estimate the pose of aerial robots in GPS denied environments. However, the pure visual system usually has poor robustness in large scale environments. This paper presents a pose estimation algorithm which fuses monocular visual and inertial data using the monocular ORB-SLAM algorithm as the visual framework. Firstly, the scale estimation algorithm is proposed to estimate the unknown scale parameter in a monocular visual framework based on IMU measurements, meanwhile the gyroscope bias of IMU and gravity direction are initialized. Then, the Extended Kalman Filter (EKF) algorithm is used to fuse the pose estimation results obtained by monocular vision system and IMU respectively. This method makes full use of the advantages of high-frequency characteristics of IMU and low drift of monocular vision system. Finally, the experiments are carried out on a small multi-rotor Unmanned Aerial Vehicle (UAV). The results show that the maximum error of position estimation is less than 1.326 meters, and the maximum error of attitude estimation is about 3.859 degrees, which verifies the accuracy, effectiveness and stability of the proposed algorithm.