Hello, could someone set up LSD SLAM&IMU sensor fusion?
Where to get the covariance matrix? Is it a fixed matrix multiplied by lsd-slam scale enough, or it is better to set pose_use_fixed_covariance=true?
I have base_link for IMU and front_cam_optical_frame for lsd_slam/pose tf frames.
I rotate lsd_slam pose in different node. Like that:
pListener->lookupTransform ( "base_link", "front_cam_optical_frame", ros::Time ( 0 ), base_link_to_camera_transform );tf::Transform pose_out = lsd_slam_pose_orientation*base_link_to_camera_transform.inverse();tf::Quaternion pose_orientation = pose_out.getRotation();
What now I should use for q_ic? [x=0, y=0, z=0, w=1] or something else?
Hello, could someone set up LSD SLAM&IMU sensor fusion?
Where to get the covariance matrix? Is it a fixed matrix multiplied by lsd-slam scale enough, or it is better to set
pose_use_fixed_covariance=true?I have base_link for IMU and front_cam_optical_frame for lsd_slam/pose tf frames. I rotate lsd_slam pose in different node. Like that:
pListener->lookupTransform ( "base_link", "front_cam_optical_frame", ros::Time ( 0 ), base_link_to_camera_transform );tf::Transform pose_out = lsd_slam_pose_orientation*base_link_to_camera_transform.inverse();tf::Quaternion pose_orientation = pose_out.getRotation();What now I should use for q_ic? [x=0, y=0, z=0, w=1] or something else?Sorry for my English. Hope for your help!