Skip to content

Latest commit

 

History

History
49 lines (26 loc) · 1.38 KB

kalibr.md

File metadata and controls

49 lines (26 loc) · 1.38 KB

Kalibr

kalibr_allan

  • rosbag record /imu0 -O imu.bag

  • rosrun bagconvert bagconvert imu.bag /imu0

  • Run the included matlab scripts to generate an allan

    deviation plot for the readings

SCRIPT_allan_matparallel.m

  • Interpret the generated charts to find noise values

SCRIPT_process_result.m

Collect calibration data

  • rosrun topic_tools throttle messages <in_topic> 4.0 [out_topic]

  • rosrun image_view image_view image:=/cam0/image_raw

  • rosbag record /cam0/image_raw /cam1/image_raw -O static.bag

  • rosrun image_view image_view image:=/cam0/image_raw

  • rosbag record /cam0/image_raw /cam1/image_raw /imu0 -O dynamic.bag

Run the calibration

动态标定的chain.yml 是 静态标定 产生的 动态标定需加 --time-calibration

  • kalibr_calibrate_cameras --models pinhole-equi pinhole-equi --topics /cam0/image_raw /cam1/image_raw --bag static.bag --target aprilgrid_6x6.yaml

  • kalibr_calibrate_imu_camera --cam chain.yaml --target aprilgrid_6x6.yaml --imu imu0.yaml --bag dynamic.bag --time-calibration

https://github.com/ethz-asl/maplab/wiki/Initial-sensor-calibration-with-Kalibr

Get config

There are many tools to generate config for popular slam algorithm.

https://github.com/ethz-asl/kalibr/tree/master/aslam_offline_calibration/kalibr/python/exporters

example:

rosrun kalibr kalibr_maplab_config --to-ncamera --label <your_label> --cam <your_cam_chain_yaml>