Real-time visual odometry for autonomous MAV navigation using RGB-D camera

In this paper, we present a visual odometry algorithm for a Micro Aerial Vehicle (MAV) navigation system using data fused from an RGB-D camera and an Inertial Measurement Unit (IMU). The Image Interpolation Algorithm (I2A) is used to calculate optic flow from the RGB-D intensity image and egomotion is recovered by combining the range data with the optic flow field Image Jacobian. An Extended Kalman Filter (EKF) is used to fuse inertial data with the egomotion recovered from the RGB-D camera. By integrating the egomotion, estimation of the velocity and position of the quadrotor is obtained in three dimensional space. A Vicon Motion Tracking System provides the position measurement which is used as ground truth for analysing the system error. Based on experiments done in an indoor environment, the accuracy of the velocity and the position estimation is evaluated.