search

koide3 / direct_visual_lidar_calibration

A toolbox for target-less LiDAR-camera calibration [ROS1/ROS2]
https://koide3.github.io/direct_visual_lidar_calibration/
702 stars 102 forks source link

Preprocessing bag point cloud does not perform dense point cloud reconstruction #106

Open oulton opened 1 month ago

oulton commented 1 month ago

docker run \ -it \ --rm \ --net host \ --gpus all \ -e DISPLAY=$DISPLAY \ -v $HOME/.Xauthority:/root/.Xauthority \ -v $bag_path:/tmp/input_bags \ -v $preprocessed_path:/tmp/preprocessed \ koide3/direct_visual_lidar_calibration:noetic \ rosrun direct_visual_lidar_calibration preprocess -av \ --camera_model plumb_bob \ --camera_intrinsic 793.12038,793.39645,642.28865,339.063046 \ --camera_distortion_coeffs -0.35688425,0.1004065,-0.00069925,0.000116,0 \ /tmp/input_bags /tmp/preprocessed

when i run preprocess node with a bag, and the car is moving. This preprocess node can perform dense point cloud reconstruction, but in docker i can not get the reconstruction cloud,it just a simple overlay of many frames.

koide3 commented 3 weeks ago

You need to add -d option to enable the dynamic point integrator.