A Fast and Robust Camera-IMU Online Calibration Method For Localization System
Authors:
Xiaowen Tao,
Pengxiang Meng,
Bing Zhu,
Jian Zhao
Abstract:
Autonomous driving has spurred the development of sensor fusion techniques, which combine data from multiple sensors to improve system performance. In particular, localization system based on sensor fusion , such as Visual Simultaneous Localization and Mapping (VSLAM), is an important component in environment perception, and is the basis of decision-making and motion control for intelligent vehicl…
▽ More
Autonomous driving has spurred the development of sensor fusion techniques, which combine data from multiple sensors to improve system performance. In particular, localization system based on sensor fusion , such as Visual Simultaneous Localization and Mapping (VSLAM), is an important component in environment perception, and is the basis of decision-making and motion control for intelligent vehicles. The accuracy of extrinsic calibration parameters between camera and IMU has significant effect on the positioning precision when performing VSLAM system. Currently, existing methods are time-consuming using complex optimization methods and sensitive to noise and outliers due to off-calibration, which can negatively impact system performance. To address these problems, this paper presents a fast and robust camera-IMU online calibration method based space coordinate transformation constraints and SVD (singular Value Decomposition) tricks. First, constraint equations are constructed based on equality of rotation and transformation matrices between camera frames and IMU coordinates at different moments. Secondly, the external parameters of the camera-IMU are solved using quaternion transformation and SVD techniques. Finally, the proposed method is validated using ROS platform, where images from the camera and velocity, acceleration, and angular velocity data from the IMU are recorded in a ROS bag file. The results showed that the proposed method can achieve robust and reliable camera-IMU online calibration parameters results with less tune consuming and less uncertainty.
△ Less
Submitted 14 May, 2023;
originally announced May 2023.
