search

rsasaki0109 / lidarslam_ros2

ROS 2 package of 3D lidar slam using ndt/gicp registration and pose-optimization
BSD 2-Clause "Simplified" License
512 stars 121 forks source link

Expected Performance with Ideal Pose #30

Open goksanisil23 opened 2 years ago

goksanisil23 commented 2 years ago

Hi,

Firstly thanks for providing this package, especially in ROS2.

My question is about usage of ideal odometry data (and the corresponding tf) for mapping purposes.

The odometry data I'm using is "ideal" in the sense that its from an RTK system with centimeter accuracy. I'm not using IMU data. However when the pose from the SLAM algorithm is updated with the pose from pointcloud matching, I can see a visible drift at the point of loop closure. So the modified_map that's generated at the end has a visible mismatch at the loop closure point (but the input odometry data is continous, without any jump) Is there some way to increase the weight of the pose information from the odometry, or some constraint parameter?

Thanks and regards

rsasaki0109 commented 2 years ago

Thank you for your interest in this package. Unfortunately, it is currently not possible to use RTK-GNSS well with this package.

If someone could provide me with sample data, I could develop a RTK-GNSS combined function.

boostthrill commented 10 months ago

I could provide you with a ROSBAG with RTK-GNSS ( /fix /heading) and the PointCloud of a Livox MID 70. Would that help you to program a combined function.

rsasaki0109 commented 10 months ago

@boostthrill Yes, the provided ROSBAG data will be very helpful for developing the combined functionality.