UB
Uploaded byUniversity of Colorado at Boulder
532 views
Lecture 07: Localization and Mapping I
This document discusses localization and mapping for robotics. It introduces topics like gyroscopes, odometry, GPS, and landmarks for localization. It discusses uncertainty models using Gaussian distributions and error propagation. Methods for belief representation are presented, including parametric single/multi hypothesis and non-parametric particle filters. Environment representations like continuous, discrete, and topological maps are described. The document provides an example of Google Maps and discusses belief representation in topological maps. It also covers multi-hypothesis belief representation, sensor data to topological maps using exact and Voronoi decompositions, and adaptive cell-size. The document assigns homework on navigation algorithms and reactive vs. deliberative planning.
More Related Content
Viewers also liked
Similar to Lecture 07: Localization and Mapping I
More from University of Colorado at Boulder
Recently uploaded
Lecture 07: Localization and Mapping I
- 1. Introduction to RoboticsLocalizationand Mapping IMarch 1, 2010
- 2. Last week’s exerciseRobotStadiumIntroductionto SVN (versioning/collaboration tool)TasksLocomotion: need to get up and runPerception: need to orient itselfCommunicate: need to share informationLocalize: need to reason about spaceDeliberate: need to plan what to do nextShare the load: 1 or 2 tasks per studentPresent your plan in 2 weeks in class – be specific
- 3.
- 4. LocalizationGyroscopeOdometryControl inputGPSLandmarksSensor inputwith different uncertainties. What is the overall uncertainty of the estimate?
- 5. Uncertainty Models: TheGaussian Distribution
- 6. Error PropagationIntuition: themore sensitive the estimated quantity is to perception error, the more this sensor should be weightedCovariance matrixRepresenting outputuncertaintiesFunction relating sensor inputto output quantitiesCovariance matrixrepresenting inputuncertainties
- 7. Differential Wheel RobotOdometry
- 8.
- 9.
- 10. Belief representationParametric, singlehypothesisParametric, multi hypothesisNon-parametric, multi hypothesis(particle filter)
- 11.
- 12. Example: Google MapsContinuous,Discrete or Topological?
- 13. Belief representation intopological maps
- 14.
- 15. From Sensor Datato Topological MapsExact Decomposition
- 16. Voronoi DecompositionPoints onlines have the same distance to neighboring obstaclesVoronoi edges correspond to the safest path
- 17.
- 18. Exercise: Navigation AlgorithmsFindthe shortest path from A to BChoose the map representationDevise an algorithm to extract path
- 19. Reactive vs. DeliberativePlanningSo farMove randomlyUse heuristics (follow wall, spiral, …)Use landmarks (infrared beacons, magnet wire)Use gradients / feedback control (Exercise 2)TodayDeliberative planningReason on abstract representation
- 20. HomeworkSection 5.6 (pages212-244)