REEF Calibrator: An Open-Source Online IMU-Camera Calibration

This tutorial presents a detailed implementation of the Multiplicative Extended Kalman Filter (MEKF) for concurrent visual-aided inertial navigation and sensor calibration. In particular, the tutorial provides the mathematics associated with implementing a MEKF to fuse Inertial Measurement Unit (IMU) measurements with known pixel image features from a monocular camera to perform global position and attitude estimation. To find the sensor calibration, i.e., the relative IMU-camera pose, the relative pose is appended to the MEKF state. The corresponding mathematical details are also included in this paper. Overall, this paper intends to provide the fundamentals for deriving process and measurement models to implement MEKF-based navigation systems.