LVI_SAM_fixed
LVI_SAM_fixed copied to clipboard
epicjung
lidar to imu extrinsic
When I studied the parameters of LIO-SAM params_lidar.yaml, I found that the author made a remark: the parameter of external reference from lidar to IMU was wrong, but it was actually the external reference from IMU to lidar. Would you please tell me whether the external reference from lidar to IMU or from IMU to lidar is in the. yaml file of your code? In addition, I don't quite understand the configuration of the external reference in the parameter file, and I haven't been able to get through KITTI yet.
@hellopopyee I used my program to run KITTI successfully using "kitti_2011_09_30_drive_0028_synced". The parameters that I used are:
from params_camera.yaml
lidar_to_cam_tx: 0.3335 lidar_to_cam_ty: -0.0048 lidar_to_cam_tz: -0.0756 lidar_to_cam_rx: -1.5638 lidar_to_cam_ry: 0.0 lidar_to_cam_rz: -1.5731
lidar_to_imu_tx: -8.086759e-01 lidar_to_imu_ty: 3.195559e-01 lidar_to_imu_tz: -7.997231e-01 lidar_to_imu_rx: -0.0008 lidar_to_imu_ry: -0.002 lidar_to_imu_rz: 0.0148
from params_lidar.yaml
extrinsicTrans: [-8.086759e-01, 3.195559e-01, -7.997231e-01] extrinsicRot: [9.999976e-01, 7.553071e-04, -2.035826e-03, -7.854027e-04, 9.998898e-01, -1.482298e-02, 2.024406e-03, 1.482454e-02, 9.998881e-01] extrinsicRPY: [1, 0, 0, 0, 1, 0, 0, 0, 1]
Please tell me if the modification does not work out for you.
Could you please show the original version of calib_velo_to_cam.txt and calib_imu_to_velo.txt? I didn't find a version that matches your extrinsics.
Sorry to bother you again,I get some extrinsics that familiar to yours,here are some differences:
lidar_to_cam_ry: -3.14156665742836072
lidar_to_imu_rx: -3.12677
lidar_to_imu_ry: -3.13957
lidar_to_imu_rz: 3.14081
I'm not sure if it is a gimbal lock problem,your extrincs work well,could you tell me how to avoid gimbal lock problem.
Sorry to bother you again,I get some extrinsics that familiar to yours,here are some differences:
lidar_to_cam_ry: -3.14156665742836072
lidar_to_imu_rx: -3.12677
lidar_to_imu_ry: -3.13957
lidar_to_imu_rz: 3.14081
I'm not sure if it is a gimbal lock problem,your extrincs work well,could you tell me how to avoid gimbal lock problem.
My calculation is the same as yours. Compared with the author's parameters, rx, ry and rz need to be processed with +/-pi, and then rx and Rz are exchanged. After converting the rotation matrix to Euler angles, why need to do that?
