MYNT-EYE-S-SDK/docs/src/slam/how_to_use_kalibr.rst

300 lines
13 KiB
ReStructuredText
Raw Normal View History

2019-03-14 05:48:41 +02:00
.. _how_to_use_kalibr:
How to calibrate MYNTEYE by kalibr
===================================
Target
------------
* Calibrate the pose relationship between left and right camera
* Calibrate the pose relationship left camera between and IMU
Preparation
------------
* **Install kalibr**:Refer to `kalibr wiki <https://github.com/ethz-asl/kalibr/wiki/installation>`_ and follow the steps to install
* **Calibration board** kalibr supports ``chessbord`` , ``circlegrid`` , ``aprilgrid`` ,choose ``aprilgrid`` here Calibration board file can be directly `download <https://github.com/ethz-asl/kalibr/wiki/downloads>`_ , Or you can also generate calibration board by Kalibr tool.
.. code-block:: bash
$ kalibr_create_target_pdf --type 'apriltag' --nx 6 --ny 6 --tsize 0.08 --tspace 0.3
View parameters' meanings by kalibr_create)target_pdf command:
.. code-block:: bash
$ kalibr_create_target_pdf --h
usage:
Example Aprilgrid:
kalibr_create_target_pdf --type apriltag --nx 6 --ny 6 --tsize 0.08 --tspace 0.3
Example Checkerboard:
kalibr_create_target_pdf --type checkerboard --nx 6 --ny 6 -csx 0.05 --csy 0.1
Generate PDFs of calibration patterns.
optional arguments:
-h, --help show this help message and exit
Output options:
output Output filename
--eps Also output an EPS file
Generic grid options:
--type GRIDTYPE The grid pattern type. ('apriltag' or 'checkerboard')
--nx N_COLS The number of tags in x direction (default: 6)
--ny N_ROWS The number of tags in y direction (default: 7)
Apriltag arguments:
--tsize TSIZE The size of one tag [m] (default: 0.08)
--tspace TAGSPACING The space between the tags in fraction of the edge size
[0..1] (default: 0.3)
--tfam TAGFAMILIY Familiy of April tags ['t16h5', 't25h9', 't25h7',
't36h11'] (default: t36h11)
Checkerboard arguments:
--csx CHESSSZX The size of one chessboard square in x direction [m]
(default: 0.05)
--csy CHESSSZY The size of one chessboard square in y direction [m]
(default: 0.05)
* **Calibrate the intrinsic IMU parameters** kalibr requires imu data to be calibrated by intrinsic parameters by default.The intrinsic parameters calibration tool uses `imu-tk <https://github.com/Kyle-ak/imu_tk.git>`_ .
* **Count imu data parameter**
* noise density
* bias random walk
Using Allan analyzing tool `imu_utils <https://github.com/gaowenliang/imu_utils>`_, We can get the characteristics of above imu data,and to format the output as ``imu.yaml``
.. code-block:: bash
#Accelerometers
accelerometer_noise_density: 0.02680146180736048 #Noise density (continuous-time)
accelerometer_random_walk: 0.0026296086159332804 #Bias random walk
#Gyroscopes
gyroscope_noise_density: 0.008882328296710996 #Noise density (continuous-time)
gyroscope_random_walk: 0.00037956578292701033 #Bias random walk
rostopic: /mynteye/imu/data_raw #the IMU ROS topic
update_rate: 200.0 #Hz (for discretization of the values above)
Calibrate the pose relationship between left and right camera
--------------------------------------------------------------
* Collect calibration images: kalibr supports the collection of the required calibration images through two ways:by ``rosbag`` or collect offline images . Using ``rosbag`` here for convenience,Reference `link <https://github.com/ethz-asl/kalibr/wiki/bag-format>`_ for collecting images.
* Method of collecting images by ``rosbag`` :fix mynteye camera,move ``aprilgrid`` calibration board in the camera field of view.
* To increase the calibration time,try to use image acquisition data with lower frame rate,kalibr recommends using ``4Hz`` frame rate,here uses ``10hz`` .
* MYNTEYE S series camera offers images at least 10Hz,You can use `topic_tools <http://wiki.ros.org/topic_tools/throttle>`_ to modify frequency,because using 10Hz requires more calibration time.
* Record ``static.bag`` : After fix the mynteye camera,start `wrapper <https://github.com/slightech/MYNT-EYE-S-SDK>`_, record the topic of the left and right images to ``static.bag`` .
.. code-block:: bash
$ source wrappers/ros/devel/setup.bash
$ roslaunch mynt_eye_ros_wrapper display.launch
$ cd ~
$ mkdir -p bag
$ cd bag
$ rosbag record -O static_10hz /mynteye/left/image_raw /mynteye/right/image_raw #recommand use 10hz,you can also use topic_tools to publish 4hz.
* kalibr calibration:
.. code-block:: bash
$ kalibr_calibrate_cameras --target aprilgrid.yaml --bag ~/bag/static_10hz.bag --models pinhole-radtan pinhole-radtan --topics /mynteye/left/image_raw /mynteye/right/image_raw
View parameters' meanings by kalibr_calibrate_cameras command:
.. code-block:: bash
$ kalibr_calibrate_cameras --h
Calibrate the intrinsics and extrinsics of a camera system with non-shared
overlapping field of view.
usage:
Example usage to calibrate a camera system with two cameras using an aprilgrid.
cam0: omnidirection model with radial-tangential distortion
cam1: pinhole model with equidistant distortion
kalibr_calibrate_cameras --models omni-radtan pinhole-equi --target aprilgrid.yaml \
--bag MYROSBAG.bag --topics /cam0/image_raw /cam1/image_raw
example aprilgrid.yaml:
target_type: 'aprilgrid'
tagCols: 6
tagRows: 6
tagSize: 0.088 #m
tagSpacing: 0.3 #percent of tagSize
optional arguments:
-h, --help show this help message and exit
--models MODELS [MODELS ...]
The camera model ['pinhole-radtan', 'pinhole-equi',
'omni-radtan', 'pinhole-fov'] to estimate
Data source:
--bag BAGFILE The bag file with the data
--topics TOPICS [TOPICS ...]
The list of image topics
--bag-from-to bag_from_to bag_from_to
Use the bag data starting from up to this time [s]
Calibration target configuration:
--target TARGETYAML Calibration target configuration as yaml file
Image synchronization:
--approx-sync MAX_DELTA_APPROXSYNC
Time tolerance for approximate image synchronization
[s] (default: 0.02)
Calibrator settings:
--qr-tol QRTOL The tolerance on the factors of the QR decomposition
(default: 0.02)
--mi-tol MITOL The tolerance on the mutual information for adding an
image. Higher means fewer images will be added. Use -1
to force all images. (default: 0.2)
--no-shuffle Do not shuffle the dataset processing order
Outlier filtering options:
--no-outliers-removal
Disable corner outlier filtering
--no-final-filtering Disable filtering after all views have been processed.
--min-views-outlier MINVIEWOUTLIER
Number of raw views to initialize statistics (default:
20)
--use-blakezisserman Enable the Blake-Zisserman m-estimator
--plot-outliers Plot the detect outliers during extraction (this could
be slow)
Output options:
--verbose Enable (really) verbose output (disables plots)
--show-extraction Show the calibration target extraction. (disables
plots)
--plot Plot during calibration (this could be slow).
--dont-show-report Do not show the report on screen after calibration.
Output the following three files after finish calibration:
* ``camchain-homezhangsbagstatic_10hz.yaml``
* ``report-cam-homezhangsbagstatic_10hz.pdf``
* ``results-cam-homezhangsbagstatic_10hz.txt``
.. tip::
If you use camera parameters in Vins,it would be better to choose the pinhole-equi model or the omni-radtan model.If you use camera parameters in Maplab,please choose pinhole-equi model
Calibrate the pose relationship between camera and IMU coordinate system
-------------------------------------------------------------------------
* **Collect calibration data**as calibrate the pose relationship of camera,Kalibr supports two ways to collect data,we still use ``rosbag`` here.
* Method of collecting image: fix ``apilgrid`` calibration board, move camera
* Make sure that the data collected is good:the brightness of the calibration board should be appropriate,too bright or too dark can't guarantee the quality of data,meanwhile do not shake too fast to avoid blurring of the image.
* Set the imu publishing frequency to 200Hz,image to 20Hz(recommended by kalibr)
* Fully motivate each axis of the imu,for example ,3 actions on each axis,then in the \"8-shaped\" motion
* Record camera and imu as ``dynamic.bag``.
.. code-block:: bash
$ roslaunch mynt_eye_ros_wrapper display.launch
$ cd bag
$ rosbag record -O dynamic /mynteye/left/image_raw /mynteye/right/image_raw /mynteye/imu/data_raw #remember set image hz to 20hz, imu hz to 200hz
* kalibr calibration:
.. code-block:: bash
$ kalibr_calibrate_imu_camera --cam camchain-homezhangsbagstatic_10hz.yaml --target aprilgrid.yaml --imu imu.yaml --time-calibration --bag ~/bag/dynamic.bag
View the parameters' meanings by kalibr_calibrate_imu_camera command
.. code-block:: bash
$ kalibr_calibrate_imu_camera --h
Calibrate the spatial and temporal parameters of an IMU to a camera chain.
usage:
Example usage to calibrate a camera system against an IMU using an aprilgrid
with temporal calibration enabled.
kalibr_calibrate_imu_camera --bag MYROSBAG.bag --cam camchain.yaml --imu imu.yaml \
--target aprilgrid.yaml --time-calibration
camchain.yaml: is the camera-system calibration output of the multiple-camera
calibratin tool (kalibr_calibrate_cameras)
example aprilgrid.yaml: | example imu.yaml: (ADIS16448)
target_type: 'aprilgrid' | accelerometer_noise_density: 0.006
tagCols: 6 | accelerometer_random_walk: 0.0002
tagRows: 6 | gyroscope_noise_density: 0.0004
tagSize: 0.088 | gyroscope_random_walk: 4.0e-06
tagSpacing: 0.3 | update_rate: 200.0
optional arguments:
-h, --help show this help message and exit
Dataset source:
--bag BAGFILE Ros bag file containing image and imu data (rostopics
specified in the yamls)
--bag-from-to bag_from_to bag_from_to
Use the bag data starting from up to this time [s]
--perform-synchronization
Perform a clock synchronization according to 'Clock
synchronization algorithms for network measurements'
by Zhang et al. (2002).
Camera system configuration:
--cams CHAIN_YAML Camera system configuration as yaml file
--recompute-camera-chain-extrinsics
Recompute the camera chain extrinsics. This option is
exclusively recommended for debugging purposes in
order to identify problems with the camera chain
extrinsics.
--reprojection-sigma REPROJECTION_SIGMA
Standard deviation of the distribution of reprojected
corner points [px]. (default: 1.0)
IMU configuration:
--imu IMU_YAMLS [IMU_YAMLS ...]
Yaml files holding the IMU noise parameters. The first
IMU will be the reference IMU.
--imu-delay-by-correlation
Estimate the delay between multiple IMUs by
correlation. By default, no temporal calibration
between IMUs will be performed.
--imu-models IMU_MODELS [IMU_MODELS ...]
The IMU models to estimate. Currently supported are
'calibrated', 'scale-misalignment' and 'scale-
misalignment-size-effect'.
Calibration target:
--target TARGET_YAML Calibration target configuration as yaml file
Optimization options:
--time-calibration Enable the temporal calibration
--max-iter MAX_ITER Max. iterations (default: 30)
--recover-covariance Recover the covariance of the design variables.
--timeoffset-padding TIMEOFFSET_PADDING
Maximum range in which the timeoffset may change
during estimation [s] (default: 0.01)
Output options:
--show-extraction Show the calibration target extraction. (disables
plots)
--extraction-stepping
Show each image during calibration target extraction
(disables plots)
--verbose Verbose output (disables plots)
--dont-show-report Do not show the report on screen after calibration.
Output the follwing 4 files after finish calibration:
* ``camchain-imucam-homezhangsbagdynamic.yaml``
* ``imu-homezhangsbagdynamatic.yaml``
* ``report-imucam-homezhangsbagdynamic.pdf``
* ``results-imucam-homezhangsbagdynamic.yaml``