survey: simultaneous localisation and mapping (slam) · 2017-06-02 · survey: simultaneous...
TRANSCRIPT
![Page 1: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/1.jpg)
Survey: SimultaneousLocalisation and Mapping
(SLAM)Ronja Güldenring
Master InformaticsProject Intellgient Robotics
University of Hamburg
![Page 2: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/2.jpg)
Introduction EKF-SLAM FastSLAM Loop Closure
Ronja Güldenring01.06.17 2
![Page 3: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/3.jpg)
Introduction EKF-SLAM FastSLAM Loop Closure
Ronja Güldenring01.06.17 3
![Page 4: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/4.jpg)
The Idea
• Basic procedures• Map Localisation
• LocalisationMap
• No prior knowledge
• SLAM‘s objective: optimum of concurrent localisation and mapping
Ronja Güldenring01.06.17 4
![Page 5: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/5.jpg)
Typical sensors
• Odometry• Encoder in wheels
• Laser scanner• 2D: scan in a plane
• 3D: scans in different planes
• Camera• 2D: rgb images
• 3D: rgb images and depth information
01.06.17 Ronja Güldenring 5
![Page 6: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/6.jpg)
Grid based mapping
• Environment is divided in grid• Value of every cell is the
probability of the occupancy ofthe grid
• Val = 0 cell is occupied
• Val = 1 cell is free
• Scan-matching via ICP• Align scans according to robot
position and already existing gridmap
01.06.17 Ronja Güldenring 6
[Hertzberg, Lingemann, Nüchter, 2012]
![Page 7: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/7.jpg)
Feature based map
• Relative position of features aresaved in map
• Extracting features of theenvironment• Re-obeservable from different
position or angles• Unique• Stationary
• Example• SIFT• Hough-Transform
01.06.17 Ronja Güldenring 7
![Page 8: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/8.jpg)
SLAM Scenario
• 𝑘: time instant
• 𝑥𝑘: robot location
• 𝑢𝑘: control vector between time 𝑘 − 1 and 𝑘.
• 𝑚: map of landmarks and theirrespective location
• 𝑧𝑘: „observation“: measurementbetween robot and landmarks
01.06.17 Ronja Güldenring 8
[Durrant-Whyte, Bailey, 2006]
![Page 9: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/9.jpg)
Full vs Online SLAM?
01.06.17 Ronja Güldenring 9
Full Slam:Estimation of the entire path:
𝑝(𝑥𝑘 , 𝑚|𝑧𝑘 , 𝑢𝑘)
Online Slam: Estimation of the most recent pose:
𝑝(𝑥𝑘 , 𝑚|𝑧𝑘 , 𝑢𝑘)
[http://ais.informatik.uni-freiburg.de/teaching/ss12/robotics/slides/12-slam.pdf]
![Page 10: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/10.jpg)
Ronja Güldenring01.06.17 10
Introduction EKF-SLAM FastSLAM Loop Closure
![Page 11: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/11.jpg)
EKF-SLAM
• First popular SLAM-approach
• Online SLAM solution
• Based on extended kalman filter
• Sketch1. Predict current system state
2. Correct current system state
3. Add new landmarks
01.06.17 Ronja Güldenring 11
ҧ𝑥 =ത𝑅𝐿𝑀
=
ത𝑅𝐿𝑀1
⋮𝐿𝑀𝑙
𝑃 =𝑃𝑅𝑅 𝑃𝑅𝐿𝑀𝑃𝐿𝑀𝑅 𝑃𝐿𝑀𝐿𝑀
(2.1)
[Solà 2014]
![Page 12: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/12.jpg)
EKF-SLAM 1. Predict current state
• Model of the physical system𝑓( )
• Robot position will be predicted• control vector u
• Perturbation n, Covariance-matrix N
• Landmarks are static
01.06.17 Ronja Güldenring 12
𝑅 ← 𝑓 𝑅, 𝑢, 𝑛, 𝑁 (2.2)𝑀 ← 𝑀 (2.3)
[Solà 2014]
![Page 13: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/13.jpg)
EKF-SLAM 2. Correct current state
• Current state will be updatedaccording to obeservationfunction ℎ( )
• Measurement• Odometry
• Landmark re-detection
• Covariance-matrix R
• 𝑂(𝐿2)
01.06.17 Ronja Güldenring 13
[Solà 2014]
![Page 14: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/14.jpg)
EKF-SLAM 3. Add new landmarks
01.06.17 Ronja Güldenring 14
ҧ𝑥 ←ҧ𝑥
𝐿𝑀𝐿+1
• New landmarks
• Dynamic system state
• System state grows with thenumber of landmarks
𝑃 ←𝑃 𝑃𝑋𝐿𝑀
𝑃𝐿𝑀𝑋 𝑃𝐿𝑀𝐿𝑀(2.4)
[Solà 2014]
![Page 15: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/15.jpg)
Ronja Güldenring01.06.17 15
Introduction EKF-SLAM FastSLAM Loop Closure
![Page 16: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/16.jpg)
Particle Filter
• Finite set of samples
• Set represents the probabilitydistribution
• Regions with high particle density high probability
1. Moving particles according tosystem model
2. Weighting particles according tomeasurements
3. Resampling particlesdeleting particles with low weightsadd new particles to the higherprobable regions
01.06.17 Ronja Güldenring 16
[Montemerlo 2003]
![Page 17: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/17.jpg)
FastSlam [Montemerlo 2003]
• Full SLAM
• Particle filter: path estimaton
• Extended kalman filter: landmark position estimation
• Low cost: O(IL), where I is the number of particles and L the numberof landmarks
Can be used in large environments with up to a millionlandmarks
01.06.17 Ronja Güldenring 17
![Page 18: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/18.jpg)
FastSlamRao-Blackwellization
• Landmark positions and the robots path are treated independently
SLAM can be factorized into L+1 estimation problems
01.06.17 Ronja Güldenring 18
Path estimator Landmark pose estimator
Robot position map observations control vector
𝑝 𝑥𝑘 , 𝑚 𝑧𝑘 , 𝑢𝑘 , 𝑛𝑘 = 𝑝 𝑥𝑘 𝑧𝑘 , 𝑢𝑘 , 𝑛𝑘 ς𝑙 𝑝 𝑚𝑙 𝑥𝑘 , 𝑧𝑘 , 𝑢𝑘 , 𝑛𝑘
![Page 19: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/19.jpg)
FastSlamPath prediction via particel filter
• Each particel 𝑠𝑘,𝑖 ∈ 𝑆𝑘represents a possible systemstate.
• Prediction• Next robot position in the path is
predicted with the system model
• The new particel set istemporary
01.06.17 Ronja Güldenring 19
[Montemerlo 2003]
![Page 20: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/20.jpg)
FastSlamLandmark location estimation via Kalman Filter
• Landmark map m will be estimated for each particle
• Extended kalman filter• Landmarks are treated independently, for every landmark a kalman filter
exists. IL kalman filters in total
• No prediction step as the position of the landmarks is assumed as static.
• Mean 𝑚𝑙𝑖: 2-element vector
• Covariance 𝑃𝑚𝑙𝑖 : 2-by-2-matrix
01.06.17 Ronja Güldenring 20
![Page 21: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/21.jpg)
FastSlamweighting and resampling
• After the landmark estimation, for each particle an importance factor𝑤𝑘𝑖 will be determined• Measurements
• Landmark position
• Temporary path
• Resampling according to importance factor• Sampling technique can be chosen variable
• Most complex part of the procedure O(LI)
01.06.17 Ronja Güldenring 21
![Page 22: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/22.jpg)
FastSlamVisualization
01.06.17 Ronja Güldenring 22
[http://ais.informatik.uni-freiburg.de/teaching/ss14/robotics/slides/14-slam-fastslamnew.pdf]
![Page 23: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/23.jpg)
FastSlamVisualization
01.06.17 Ronja Güldenring 23
[http://ais.informatik.uni-freiburg.de/teaching/ss14/robotics/slides/14-slam-fastslamnew.pdf]
![Page 24: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/24.jpg)
FastSlamVisualization
01.06.17 Ronja Güldenring 24
[http://ais.informatik.uni-freiburg.de/teaching/ss14/robotics/slides/14-slam-fastslamnew.pdf]
![Page 25: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/25.jpg)
Quick Overview
EKF-SLAM
• Online SLAM
• 𝑂(𝐿2)
for smaller areas
FastSLAM
• Full SLAM
• 𝑂(𝐼𝐿) / 𝑂(𝐼 log(𝐿))
usable in larger areas
01.06.17 Ronja Güldenring 25
L: number of landmarks, I: number of particles
![Page 26: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/26.jpg)
Ronja Güldenring01.06.17 26
Introduction EKF-SLAM FastSLAM Loop Closure
![Page 27: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/27.jpg)
Loop Closure
• Recognizing an already mapped area after having explored otherregions
• Uncertainties will be reduced
• Robots pose/path and landmarks, all landmark position will becorrected and get more accurate
• Challenge: Reliable data associationwrong loop closure lead to big errors
01.06.17 Ronja Güldenring 27
![Page 28: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/28.jpg)
Loop ClosureExample
01.06.17 Ronja Güldenring 28
Before closing loop After closing loop
[Montemerlo 2003]
![Page 29: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/29.jpg)
01.06.17 Ronja Güldenring 29
[https://www.youtube.com/watch?v=qpTS7kg9J3A]
![Page 30: Survey: Simultaneous Localisation and Mapping (SLAM) · 2017-06-02 · Survey: Simultaneous Localisation and Mapping (SLAM) Ronja Güldenring Master Informatics Project Intellgient](https://reader033.vdocuments.net/reader033/viewer/2022041700/5e412fc31adf7f1e61381a87/html5/thumbnails/30.jpg)
References
[Durrant-Whyte, Bailey, 2006] Hugh Durrant-Whyte, Tim Bailey. Simultaneous Localization and Mapping: Part 1. IEEE, 2006.[Hidalgo, Bräunl, 2015] Franco Hodalgo, Thomas Bräunl. Review ofUnderwater SLAM Techniques. IEEE, 2015.[Hertzberg, Lingemann, Nüchter, 2012] J. Hertzberg, K. Lingemann, A. Nüchter. Mobile Roboter – Eine Einführung aus der Sicht der Informatik. 1. Auflage. Springer-Verlag, 2012.[Montemerlo 2003] M. Montemerlo. FastSLAM: A Factored Soltution to theSimultaneous Localization and Mapping Problem With Unknown Data Association. 2003.[Solà 2014] J. Solà. Simultaneous localization and mapping with extendedKalman filter – A very quick guide… with Matlab code!. 2014.
Ronja Güldenring01.06.17 30