A high-precision calibration approach for Camera-IMU pose parameters with adaptive constraints of multiple error equations

In the calibration of the pose parameters of a camera and inertial measurement unit (Camera-IMU), the camera depth information is unreliable due to the uneven spatial distribution of calibration points, because the calibration points have random errors due to the IMU drift and the inadequate robustness of stereovision and because the Camera-IMU pose parameters lack self-adaptation. This paper proposes a high-precision calibration approach for Camera-IMU pose parameters with adaptive constraints of multiple error equations (adaptive constraint calibration approach, ACCA). The approach calibrates pose parameters of the Camera-IMU jointly via error equations, such as lens distortion correction, camera parallax correction and error compensation of the inertial sensor. The experimental results show that the calibration approach for Camera-IMU pose parameters with adaptive constraints of multiple error equations improves the measurement accuracy by 84.0% and can effectively suppress IMU drift with good robustness.