BabelCalib. A Universal Approach to Calibrating Central Cameras

Yaroslava Lochman, Kostiantyn Liepieshov, Jianhui Chen, Michal Perdoch, Christopher Zach, and James Pritts
ICCV 2021

code · data · paper · supplemental · poster · slides · presentation

BabelCalib is a calibration framework that can estimate camera models for all types of central projection cameras. Calibration is robust and fully automatic. BabelCalib provides models for pinhole cameras with additive distortion as well as omni-directional cameras and catadioptric rigs. The supported camera models are listed under the solvers directory. BabelCalib supports calibration targets made of a collection of calibration boards, i.e., multiple planar targets. The method is agnostic to the pattern type on the calibration boards. It is robust to inaccurately localized corners, outlying detections and occluded targets.

Abstract

Existing calibration methods occasionally fail for large field-of-view cameras due to the non-linearity of the underlying problem and the lack of good initial values for all parameters of the used camera model. This might occur because a simpler projection model is assumed in an initial step, or a poor initial guess for the internal parameters is pre-defined. A lot of the difficulties of general camera calibration lie in the use of a forward projection model. We side-step these challenges by first proposing a solver to calibrate the parameters in terms of a back-projection model and then regress the parameters for a target forward model. These steps are incorporated in a robust estimation framework to cope with outlying detections. Extensive experiments demonstrate that our approach is very reliable and returns the most accurate calibration parameters as measured on the downstream task of absolute pose estimation on test sets.

Data

We provide datasets in a Deltille format. The Deltille detector is a robust deltille and checkerboard detector. It comes with detector library, example detector code, and MATLAB bindings. BabelCalib provides functions for calibration and evaluation using the Deltille software’s outputs. Calibration from Deltille detections requires format conversion which is peformed by import_ODT. Please see a complete calibration example from Deltille data for planar targets and cube targets.

OV dataset

Cameras: 16 lenses ranging from 90° to 180° FOV

• 2012-A0 (single-plane target)
• 3136-H0 (single-plane target)
• 5501-C4 (single-plane target)
• 130108MP (single-plane target)
• ov00—ov07 (corner target)
• ov00—ov03 (cube target)

Calibration targets: 3 types of checkerboards containing AprilTags:

• single-plane — one planar board that has thefollowing configurations:
  • cols: 22
  • rows: 22
  • squareSize: 40 mm
• corner — 3 planar boards, each board has thefollowing configurations:
  • cols: 18
  • rows: 18
  • squareSize: 60 mm
• cube — 4 planar boards, each board has the followingconfigurations:
  • cols: 14
  • rows: 14
  • squareSize: 7.5 mm

Kalibr dataset

Cameras:

• 6 lenses ranging from 126° to 195° FOV
  • BF2M2020S23
  • BF5M13720
  • BM2820
  • BM4018S118
  • EUROC
  • GOPRO
• 1 ENTANIYA-M12 lens with 250° FOV
• 1 camera from TUM-VI with 195° FOV

Calibration target: Aprilgrid with the following configurations:

• tagCols: 6
• tagRows: 6
• tagSize: 88 mm
• tagSpacing: 0.3 # -> space in mm is 0.3 * 88 mm = 26.4 mm

OCamCalib dataset

Cameras:

• 5 fisheye cameras
  • Fisheye1
  • Fisheye2
  • Fisheye190deg
  • GOPR
  • Ladybug
• 4 catadioptric systems
  • KaidanOmni
  • MiniOmni
  • Omni
  • VMRImage

Calibration target: checkerboards of differenet (unknown) sizes.

UZH dataset

Cameras:

• 4 fisheye cameras “DAVIS”
  • indoor_45
  • indoor_forward
  • outdoor_45
  • outdoor_forward
• 4 fisheye cameras “Snapdragon”
  • indoor_45
  • indoor_forward
  • outdoor_45
  • outdoor_forward

Calibration target: Aprilgrid with the following configurations:

• tagCols: 5
• tagRows: 4
• tagSize: 75 mm
• tagSpacing: 0.2 # ~> space in mm is 0.2 * 75 mm = 15 mm

Citation

If you find BabelCalib and/or OV dataset useful, consider citing:

@InProceedings{Lochman-ICCV21,
    title     = {BabelCalib: A Universal Approach to Calibrating Central Cameras},
    author    = {Lochman, Yaroslava and Liepieshov, Kostiantyn and Chen, Jianhui and Perdoch, Michal and Zach, Christopher and Pritts, James},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
    year      = {2021},
    pages     = {15253-15262}
}