A Robust Graph-based Framework for Building Precise Maps from Laser - - PowerPoint PPT Presentation

▶
a robust graph based framework for building precise maps
SMART_READER_LITE
LIVE PREVIEW

A Robust Graph-based Framework for Building Precise Maps from Laser - - PowerPoint PPT Presentation

Jul 20, 2023 116 likes •321 views

A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans Marian Himstedt Sabrina Keil Sven Hellbach Hans-Joachim Bhme Artificial Intelligence Lab University of Applied Sciences Dresden Himstedt, Hellbach, Bhme

slide-1
SLIDE 1

Himstedt, Hellbach, Böhme MiWoCI 2012: Learning local environmental features with NMF

  • 02. – 04. July 2012

A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans

Marian Himstedt Sabrina Keil Sven Hellbach Hans-Joachim Böhme

Artificial Intelligence Lab University of Applied Sciences Dresden

slide-2
SLIDE 2

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

  • Preparing mobile robots for industrial environments:
  • Requires precise position estimates
  • Setting of artificial markers is inconvenient
  • Localization quality depends on:
  • Accuracy of sensors used
  • Computational power
  • Accuracy and resolution of prior map
  • Requirements of a SLAM framework:
  • Robust in the presence of repetitive structures
  • High scalabilty for application in large scale environments
  • High precision of final map

Introduction

2

slide-3
SLIDE 3

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

  • State-of-the-art graph optimization based methods used
  • Use of feature based SLAM
  • Scales well with larger map sizes
  • Allows efficient map matching
  • Perceptual aliasing poses a challenge
  • Limited observation space of 2D range scans
  • Industrial environments: high number of repetitive structures
  • Decouple pose and map optimization
  • Estimate pose graph topology first
  • Map optimization based on correct pose graph

Overview

3

slide-4
SLIDE 4

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Framework Overview

4

slide-5
SLIDE 5

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Front-End: Feature Extraction

5

  • Extraction of FLIRT interest points (Tipaldi et al., ICRA ’10)

Beta grid describing local surroundings Features extracted from smoothed range readings

slide-6
SLIDE 6

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Front-End: Feature Extraction

6

Example: Features detected, colors indicating scale

slide-7
SLIDE 7

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

  • Match features of reference & observed scans
  • RANSAC based outlier rejection
  • Estimation of rigid transformation of feature sets
  • Minimize point wise reprojected error

Front-End: Loop Closure Detections

7

Matching feature sets; blue: inliers, red: outliers

slide-8
SLIDE 8

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

  • Switchable Constraints (Sünderhauf et al.,

IROS ’11):

  • Loop closure incorporation is subject to optimization
  • Loop closure constraints can be “switched off”
  • Joint optimization of odometry & loop closure

constraints

  • Switch priors: Confidence provided by front-

end

  • Different switch functions possible
  • Research Lines:
  • Latif et al.: Robust Loop Closing over time (RSS ’12)
  • Agarwal et al.: Max Mixture (RSS ’12), Dynamic

Covariance Scaling (ICRA ’13)

Back-End: Pose Graph Optimization

8

slide-9
SLIDE 9

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

  • Based on Sparse Surface Adjustment (Ruhnke et al.,

ICRA ’11)

  • Assumption: Given pose graph is topologically

consistent

  • Advanced Sensor Model incorporates:
  • Incident angle w.r.t. surfaces
  • Conic shape of beam
  • Data Association: laser beams are assigned surface

patches

  • Jointly optimize robot poses and laser

measurements (range & direction)

Back-End: Map Optimization

9

surface d

r ˆ n α α dk

xi xj µni, Σni

  • µnj, Σnj
  • ∆µij

tangent error normal error

Image courtesy by Ruhnke et al.

slide-10
SLIDE 10

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Experiments

10

SCITOS G5 operating in a warehouse

slide-11
SLIDE 11

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Experiments: Robust optimization

11

Initial Pose Graph

slide-12
SLIDE 12

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Experiments: Robust optimization

12

Nonrobust optimization

slide-13
SLIDE 13

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Experiments: Robust optimization

13

Switchable Constraints

slide-14
SLIDE 14
slide-15
SLIDE 15

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Experiments: Mapping Results (I)

15

No Optimization Optimization using SSA

slide-16
SLIDE 16

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Experiments: Mapping Results (II)

16

GT [mm] ∆[mm] L1 1490.0 49.24 L2 762.0 47.05 L3 791.0 40.82 L4 650.0 8.34 L5 892.0 39.48 L6 1206.0 20.04 µ(L)

  • 34.16

Var(L)

  • 26.64

Table I

slide-17
SLIDE 17

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Experiments: Mapping Results (II)

17

No Optimization Optimization using SSA

slide-18
SLIDE 18

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

Experiments: Mapping Results (II)

18

SCITOS G5 operating in a warehouse

slide-19
SLIDE 19

Himstedt, Keil, Hellbach, Böhme A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans 10th May 2013

  • Framework is able to generate accurate maps
  • Front-end: FLIRT allows efficient place recognition
  • Pose graph: Robust optimization necessary
  • SSA: Promising results, particularly for large surfaces
  • Finding the right representation for localization
  • Low resolution global occupancy grid map
  • High resolution submaps in workspaces
  • Coping with dynamic change occuring over time
  • Dynamic Occupancy Grid Maps (Meyer-Delius et al., AAAI ’12)

Conclusion & Future Work

19