Relative Pose Of IMU-Camera Calibration Method And It's Application

IMU-Camera relative position,attitude calibration and application refers to the coupling of the IMU(inertial measurement unit)on the camera to find the relationship between the IMU coordinate system and the Camera coordinate system,so as to realize the application of IMU feedback data in electronic image stabilization.To obtain the relative pose between two sensors,we can effectively use one of the sensor information to the data feedback of the other sensor.Traditional electronic image stabilization technology process images through their feature information,which is difficult to realize when the video image sequence is not stable,blurred or deficient.Therefore,this paper proposes a method based on IMU to achieve electronic image stabilization.This method can deal with calculation errors better when the image feature information is missing or less.First,we try to fix the position of the IMU gravity center and the camera optical center together in order to decrease position translation deviation.The fusions of data from gyroscope,accelerometer and magnetic sensor help us to implement the calibration of relative position respectively,we'll then find the rotation and translation relationship between the coordinate system of IMU and camera using the quaternion method.Second,in order to synchronize the calibration time of the two sensors,we make the frequency of the two sensors consistent by using the newton interpolation and use the least square method to match the motion trajectory of the camera with the IMU to find the minimum error,so as to obtain the trigger time of both the Camera and IMU.Finally,we use the IMU fusion data to obtain the rotation matrix that will be used to calculate the homography matrix in order to obtain the coordinate translation relationship before and after the image stabilization.This paper compares and analyzes different scenes using different algorithms.The experimental results show that the electronic image stabilization method based on IMU-Camera calibration can effectively overcome the limitations of the scene and has higher robustness and application value.