The role of Planes and Edges in Lidar-Inertial Integration Teresa - - PowerPoint PPT Presentation

▶
the role of planes and edges in lidar inertial
SMART_READER_LITE
LIVE PREVIEW

The role of Planes and Edges in Lidar-Inertial Integration Teresa - - PowerPoint PPT Presentation

Dec 06, 2023 201 likes •443 views

The role of Planes and Edges in Lidar-Inertial Integration Teresa Vidal-Calleja Montreal, May 2019 cas.uts.edu.au LIDAR - INERTIAL 3D- Lidar IMU (300kHz points) Lidar Depth (10Hz) IMU Acceleration, velocity (100Hz) Sparse

slide-1
SLIDE 1

The role of Planes and Edges in Lidar-Inertial Integration Teresa Vidal-Calleja

cas.uts.edu.au

Montreal, May 2019

slide-2
SLIDE 2

LIDAR - INERTIAL

  • Lidar → Depth (10Hz)
  • IMU →Acceleration, velocity (100Hz)
  • Sparse point cloud (low vertical resolution)
  • Motion during lidar sweep → Motion distortion

cas.uts.edu.au |

IMU 3D- Lidar

Raw lidar - sparse Motion distortion (300kHz points)

slide-3
SLIDE 3

APPROACH

A probabilistic framework for localisation, mapping, and extrinsic calibration based

  • n a 3D-lidar and a 6-DoF-IMU

Key ideas

  • Point-to-map plane minimisation from frame to target for simultaneous

localisation and calibration

  • Point-to-edge/plane minimisation from frame to frame for simultaneous

localisation, mapping and autocalibration

  • Upsampled Preintegrated Measurements for point transformation

> Motion distortion correction without explicit motion model

  • Probabilistic full batch on-manifold optimisation

cas.uts.edu.au

slide-4
SLIDE 4

APPROACH OVERVIEW

cas.uts.edu.au cas.uts.edu.au

slide-5
SLIDE 5

ROLE OF PLANE AND EDGE FEATURES

Dense Lidar data:

  • ICP (and variations)
  • Surfels
  • 3D NDT, etc

Sparse Lidar data [1]:

  • Keypoints that belong to edges and lines
  • Channel by channel extraction
  • Point to plane and line formulation for estimation

cas.uts.edu.au

[1] Ji Zhang and Sanjiv Singh, “LOAM: Lidar Odometry and Mapping in Real-time” Robotics: Science and Systems Conference, 2014.

slide-6
SLIDE 6

ROLE OF PLANE AND EDGE FEATURES

cas.uts.edu.au

FEATURE EXTRACTION and MATCHING Features

  • Points with high score of being planar or corner features
  • Channel per channel

Matching

  • Point reprojection to beginning of lidar frame
  • Dependent on current state and IMU interpolation to

correct for motion distortion

  • Nearest Neighbour search

Planar points Corners

slide-7
SLIDE 7

ROLE OF PLANE AND EDGE FEATURES

FORMULATION Point to explicit plane Point to implicit plane Point to implicit line

cas.uts.edu.au

calibration SLAM

slide-8
SLIDE 8

ROLE OF PLANE AND EDGE FEATURES

FORMULATION Maximum Likelihood Estimation with as the Lidar Factor point-to-map plane/ plane / line

slide-9
SLIDE 9

APPROACH OVERVIEW

cas.uts.edu.au

slide-10
SLIDE 10

IMU PRE-INTEGRATION

cas.uts.edu.au

Preintegrated measurements [2]

  • Independent of the initial conditions and processed before optimisation

Example on velocities Integration Preintegration

Initial conditions Preintegrated measurements (only depends on IMU readings) ( f = accelerometer readings) Initial conditions + gyroscope integration Initial conditions

[2] T. Lupton and S. Sukkarieh, “Visual-inertial-aided navigation for high-dynamic motion in built environments without initial conditions” IEEE Transactions

  • n Robotics, vol. 28, no. 1, pp. 61–76, 2012
slide-11
SLIDE 11

UPSAMPLED PREINTEGRATED IMU

cas.uts.edu.au

IMU interpolation

  • Gaussian Process Regression
  • Continuous measurement

representation

  • Preintegration over inferred

measurements

  • No explicit motion model

IMU Measurements

IMU timestamps Lidar timestamps

Time

IMU raw measurements IMU inferred measurements

slide-12
SLIDE 12

LIDAR-IMU CALIBRATION

cas.uts.edu.au

  • Estimation of IMU-Lidar 3D transformation
  • Simple calibration target (set of planes)
  • Point-to-map plane minimisation from frame to target for simultaneous

localisation and calibration

IMU 3D- Lidar

Rc, pc ?

ICRA’18

slide-13
SLIDE 13

LIDAR-IMU CALIBRATION

cas.uts.edu.au

  • First static scan → Map creation (set of plane,

e.g. a room corner)

IMU 3D- Lidar

Rc, pc ?

ICRA’18

slide-14
SLIDE 14

CALIBRATION OVERVIEW

cas.uts.edu.au

Calibration target 3+ planes (E.g. room corner) Map creation First sweep static Plane equations System motion Point to map-plane correspondence Simultaneous calibration and localisation Lidar points as independent measurements

  • t0
  • t1
  • t2

ICRA’18

slide-15
SLIDE 15

SIMULTANEOUS CALIBRATION AND LOCALISATION

Nodes (state to estimate) IMU poses Lidar scan timestamps Calibration parameters Factors (constraints) IMU factors Lidar factors

cas.uts.edu.au |

t0 I0 t1 I1 t2 I2 tM IM tm Im Scan 0 Scan 1 ICRA’18

slide-16
SLIDE 16

CALIBRATION RESULTS

cas.uts.edu.au

Comparison with chained calibration with camera

Lidar - camera reprojection Chained lidar – IMU/IMU - camera reprojection

IMU Lidar RGB-D camera ICRA’18

slide-17
SLIDE 17

IN2LAMA

INERTIAL - LIDAR LOCALISATION AND MAPPING

cas.uts.edu.au |

  • Point-to-plane and point-to-line
  • On-manifold full batch optimisation
  • Lidar/bias/time-shift factors

ICRA’19

slide-18
SLIDE 18

IN2LAMA

INERTIAL - LIDAR LOCALISATION AND MAPPING

cas.uts.edu.au |

ICRA’19

MLE Cost Function

slide-19
SLIDE 19

IN2LAAMA - INERTIAL - LIDAR

LOCALISATION AUTOCALIBRATION AND MAPPING

cas.uts.edu.au

  • Point-to-plane and point-to-line distance minimisation
  • On-manifold full batch optimisation
  • Lidar/bias/time-shift/IMU factors
  • Autocalibration

archive 1905.09517

slide-20
SLIDE 20

IN2LAAMA - INERTIAL - LIDAR

LOCALISATION AUTOCALIBRATION AND MAPPING

cas.uts.edu.au

MLE Cost Function

archive 1905.09517

slide-21
SLIDE 21

FRONT END

  • Planar and edge features
  • Frame-to-frame matching
  • Tight front end/back end interaction
  • Scan greater than 360 deg
  • Back and forth data association

cas.uts.edu.au

slide-22
SLIDE 22

RESULTS

cas.uts.edu.au

LOAM IN2LAMA IN2LAAMA

slide-23
SLIDE 23

ACKNOWLEDGMENTS

cas.uts.edu.au

Cedric Le Gentil

video link