A joint calibration method of IMU-visible light camera for unmanned ship

When the unmanned ship performs tasks at sea, they typically rely on multi-sensor data fusion for environmental perception. Given the complexity of the marine environment, particularly under dynamic conditions, the calibration status among sensors on unmanned ship directly impacts the effectiveness of data fusion, subsequently affecting the navigation and positioning capabilities of unmanned ship. In fact, the unmanned ship is usually equipped with a wide range of sensors, such as LiDAR, navigation radar, IMU, visible light camera, infrared camera and so on. This paper primarily focuses on IMU-visible light cameras and introduces an external parameter joint calibration method to address the challenges of multi-sensor calibration in complex environments for unmanned ship. Initially, the pose of the visible camera is determined by calculating the visual odometer based on visible light observations, which is then registered with the GNSS/INS combined pose. Ultimately, the external parameters of the IMU-visible light camera are optimized using the least squares method. Experimental results demonstrate that this method efficiently and effectively achieves external parameter calibration for IMU-visible light cameras, significantly enhancing the navigation and positioning capabilities of unmanned ship.