#### DMCA

## Robotic mapping: A survey (2002)

### Cached

### Download Links

- [www.sci.brooklyn.cuny.edu]
- [www.sci.brooklyn.cuny.edu]
- [www.cs.uu.nl]
- [www.ai.rug.nl]
- [robots.stanford.edu]
- [cogvis.nada.kth.se]
- [www.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [www-2.cs.cmu.edu]
- [www.cs.cmu.edu]
- [robotics.stanford.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [www.cs.uml.edu]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [robotics.stanford.edu]
- [www.forengineers.net]
- [www.societyofrobots.com]
- [www8.cs.umu.se]
- [reports-archive.adm.cs.cmu.edu]
- [reports-archive.adm.cs.cmu.edu]
- [www.cs.uml.edu]
- [www8.cs.umu.se]
- [reports-archive.adm.cs.cmu.edu]
- [robotics.stanford.edu]
- [www.cs.cmu.edu]

Venue: | EXPLORING ARTIFICIAL INTELLIGENCE IN THE NEW MILLENIUM |

Citations: | 368 - 6 self |

### Citations

11970 | Maximum likelihood from incomplete data via the em algorithm
- Dempster, Laird, et al.
- 1977
(Show Context)
Citation Context ... exist that represent environments by large numbers of raw range measurements [60]. An alternative family of algorithms [23, 86, 87, 97, 101] is based on Dempster’s expectation maximization algorith=-=m [24, 65]-=-. These approaches specifically address the correspondence problem in mapping, which is the problem of determining whether sensor measurement recorded at different points in time correspond to the sam... |

5891 | A tutorial on hidden Markov models and selected applications in speech recognition,”
- Rabiner
- 1989
(Show Context)
Citation Context ... In the field of robot mapping, the single dominating scheme for integrating such temporal data is known as Bayes filter [45], which is highly related to Kalman filters [49, 64], hidden Markov models =-=[78]-=-, dynamic Bayes networks [82], and partially observable Markov decision processes [48, 59, 66, 93]. The Bayes filter extends Bayes rule to temporal estimation problems. It is a recursive estimator for... |

5448 |
Artificial Intelligence — A Modern Approach
- Russell, Norvig
- 1995
(Show Context)
Citation Context ...g, the single dominating scheme for integrating such temporal data is known as Bayes filter [45], which is highly related to Kalman filters [49, 64], hidden Markov models [78], dynamic Bayes networks =-=[82]-=-, and partially observable Markov decision processes [48, 59, 66, 93]. The Bayes filter extends Bayes rule to temporal estimation problems. It is a recursive estimator for computing a sequence of post... |

3855 |
A new approach to linear filtering and prediction problems
- Kalman
- 1960
(Show Context)
Citation Context ...eflect the actual robot motion. In the field of robot mapping, the single dominating scheme for integrating such temporal data is known as Bayes filter [45], which is highly related to Kalman filters =-=[49, 64]-=-, hidden Markov models [78], dynamic Bayes networks [82], and partially observable Markov decision processes [48, 59, 66, 93]. The Bayes filter extends Bayes rule to temporal estimation problems. It i... |

1468 |
The EM Algorithm and Extensions
- McLachlan, Krishnan
- 1997
(Show Context)
Citation Context ... exist that represent environments by large numbers of raw range measurements [60]. An alternative family of algorithms [23, 86, 87, 97, 101] is based on Dempster’s expectation maximization algorith=-=m [24, 65]-=-. These approaches specifically address the correspondence problem in mapping, which is the problem of determining whether sensor measurement recorded at different points in time correspond to the sam... |

1144 | An Introduction to the Kalman Filter
- Welch, Bishop
- 1995
(Show Context)
Citation Context ...mate shows that dependencies are local, an effect that is exploited by some algorithms that build local maps. in Equation (7) can be calculated conveniently using the standard Kalman filter equations =-=[49, 64, 103]: µ ′ t−1 = µt−1 + But Σ ′ t−1 = Σt��-=-�1 + Σcontrol Kt = Σ ′ t−1C T (CΣ ′ t−1C T + Σmeasure) −1 µt = µ ′ t−1 + Kt(o − Cµ ′ t−1) Σt = (I − KtC)Σ ′ t−1 (12) As the reader may verify, these equations are eq... |

1095 | Planning and acting in partially observable stochastic domains.
- Kaelbling, Littman, et al.
- 1998
(Show Context)
Citation Context ...oral data is known as Bayes filter [45], which is highly related to Kalman filters [49, 64], hidden Markov models [78], dynamic Bayes networks [82], and partially observable Markov decision processes =-=[48, 59, 66, 93]-=-. The Bayes filter extends Bayes rule to temporal estimation problems. It is a recursive estimator for computing a sequence of posterior probability distributions over quantities that cannot be observ... |

1074 | Modeling and rendering architecture from photographs: A hybrid geometry- and image-based approach,
- Debevec, Taylor, et al.
- 1996
(Show Context)
Citation Context ...], of furniture and other objects that move [6, 83]. Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection that is still somewhat underexploite=-=d [1, 3, 5, 16, 21, 43, 88]. -=-Robot exploration in the context of mapping has also been studied extensively. Today’s approaches are usually greedy, that is, they chose control by greedily maximizing information gain [12, 17, 89,... |

949 |
Tracking and Data Association,
- Bar-Shalom, Fortmann
- 1988
(Show Context)
Citation Context ...ion. The effect is an increase in the location uncertainty of each landmark over time, which subsequently may be counteracted by sensing. Such algorithms are popular in the target tracking literature =-=[4, 68, 85]-=-. Similarly, occupancy grid maps may accommodate certain types of motions, by decaying occupancy over time as discussed in [106]. Extensions of occupancy grids exist that can detect frequently-changin... |

838 | Monte carlo localization for mobile robots
- Dellaert, Fox, et al.
- 1999
(Show Context)
Citation Context ...ows a sequence of map estimation steps using this approach, for the same data that were used to generate Figure 6. The pose posterior estimate p(st | z t , u t ) is implemented using particle filters =-=[22, 28, 57, 77]-=-, which is a version of the Bayes filter that represents posteriors by samples. The samples can be seen in all three diagrams in Figure 7. When the robot traverses a cyclic environment, it uses the sa... |

778 | A new extension of the kalman filter to nonlinear systems,
- Julier, Uhlmann, et al.
- 1997
(Show Context)
Citation Context ...earities, Kalman filters approximate the robot motion model using a linear function obtained via Taylor series expansion. The resulting Kalman filter is known as extended Kalman filter [64] (see also =-=[47]-=-), and single motion commands are often approximated by a series of much smaller motion segments, to account for nonlinearities. For most robotic vehicles, such an approximation works well. The result... |

775 | Filtering via simulation : auxiliary particle filters.
- Pitt, Shephard
- 1999
(Show Context)
Citation Context ...ows a sequence of map estimation steps using this approach, for the same data that were used to generate Figure 6. The pose posterior estimate p(st | z t , u t ) is implemented using particle filters =-=[22, 28, 57, 77]-=-, which is a version of the Bayes filter that represents posteriors by samples. The samples can be seen in all three diagrams in Figure 7. When the robot traverses a cyclic environment, it uses the sa... |

663 | Sequential Monte Carlo methods for dynamical system,
- Liu, Chen
- 1998
(Show Context)
Citation Context ...ows a sequence of map estimation steps using this approach, for the same data that were used to generate Figure 6. The pose posterior estimate p(st | z t , u t ) is implemented using particle filters =-=[22, 28, 57, 77]-=-, which is a version of the Bayes filter that represents posteriors by samples. The samples can be seen in all three diagrams in Figure 7. When the robot traverses a cyclic environment, it uses the sa... |

599 | FastSLAM: A Factored Solution to the Simultaneous Localization and Mapping Problem,
- Montemerlo, Thrun, et al.
- 2002
(Show Context)
Citation Context ...as led to a range of extensions that can handle larger number of features, by decomposing the problem into multiple smaller ones [56, 38]. Some techniques, such as the FastSLAM algorithm described in =-=[67]-=-, promise a reduction to O(log K) complexity for certain situation, by using non-classical statistical sampling techniques for robot path estimation [29, 71] along with efficient tree representations.... |

531 | Globally consistent range scan alignment for environment mapping
- Lu, Milios
- 1997
(Show Context)
Citation Context ...ng maps usually describe the location of landmarks, or significant features in the environment, although recent extensions exist that represent environments by large numbers of raw range measurements =-=[60]. -=-An alternative family of algorithms [23, 86, 87, 97, 101] is based on Dempster’s expectation maximization algorithm [24, 65]. These approaches specifically address the correspondence problem in mapp... |

520 |
Estimating uncertain spatial relationships in robotics,” in Autonomous Robot Vehicles,
- Smith, Self, et al.
- 1988
(Show Context)
Citation Context ...nt approaches to date generate world-centric maps. Since the 1990s, the field of robot mapping has been dominated by probabilistic techniques. A series of seminal papers by Smith, Self, and Cheeseman =-=[91, 92]-=- introduced a powerful statistical framework for simultaneously solving the mapping problem and the induced problem of localizing the robot relative to its growing map. Since then, robotic mapping has... |

484 | The vector field histogram - fast obstacle avoidance for mobile robots,’’
- Borenstein, Koren
- 1991
(Show Context)
Citation Context ...lgorithm [31, 32, 69], which represents maps by fine-grained grids that model the occupied and free space of the environment. This approach has been used in a great number of robotic systems, such as =-=[8, 9, 10, 42, 83, 98, 106, 107]-=-. An alternative metric mapping algorithm was proposed by Chatila and Laumond [15], using sets of polyhedra to describe the geometry of environments. Examples of topological approaches include the wor... |

483 | Probabilistic Approach to Concurrent Mapping and Localization for Mobile Robots,
- Thrun, Fox, et al.
- 1997
(Show Context)
Citation Context ...o its growing map. Since then, robotic mapping has commonly been referred to as SLAM or CML, which is short for simultaneous localization and mapping [25, 30], and concurrent mapping and localization =-=[56, 101]-=-, respectively. One family of probabilistic approaches employ Kalman filters to estimate the map and the robot location [14, 20, 27, 38, 55, 73, 104]. The resulting maps usually describe the location ... |

470 | A robot exploration and mapping strategy based on a semantic hierarchy of spatial representations
- Kuipers, Byun
- 1991
(Show Context)
Citation Context ...ping algorithm was proposed by Chatila and Laumond [15], using sets of polyhedra to describe the geometry of environments. Examples of topological approaches include the work by Matarić [62], Kuipers=-= [53]-=- and many others [17, 18, 34, 52, 76, 87, 86, 102, 105, 108]. Topological maps represent environments as a list of significant places that are connected via arcs. Arcs are usually annotated with infor... |

420 |
The Optimal Control of Partially Observable Markov Processes
- Sondik
- 1971
(Show Context)
Citation Context ...oral data is known as Bayes filter [45], which is highly related to Kalman filters [49, 64], hidden Markov models [78], dynamic Bayes networks [82], and partially observable Markov decision processes =-=[48, 59, 66, 93]-=-. The Bayes filter extends Bayes rule to temporal estimation problems. It is a recursive estimator for computing a sequence of posterior probability distributions over quantities that cannot be observ... |

405 |
Sensor fusion in certainty grids for mobile robots”,
- Moravec
- 1988
(Show Context)
Citation Context ... environment, whereas topological maps describe the connectivity of different places. An early representative of the former approach was Elfes and Moravec’s important occupancy grid mapping algorith=-=m [31, 32, 69]-=-, which represents maps by fine-grained grids that model the occupied and free space of the environment. This approach has been used in a great number of robotic systems, such as [8, 9, 10, 42, 83, 98... |

377 |
On the representation and estimation of spatial uncertainty. The international journal of Robotics
- Smith, Cheeseman
- 1986
(Show Context)
Citation Context ...nt approaches to date generate world-centric maps. Since the 1990s, the field of robot mapping has been dominated by probabilistic techniques. A series of seminal papers by Smith, Self, and Cheeseman =-=[91, 92]-=- introduced a powerful statistical framework for simultaneously solving the mapping problem and the induced problem of localizing the robot relative to its growing map. Since then, robotic mapping has... |

361 | Markov localization for mobile robots in dynamic environments.
- Fox, Burgard, et al.
- 1999
(Show Context)
Citation Context ...a map would be quite simple. Conversely, if we already had a map of the environment, there exist computationally elegant and efficient algorithms for determining the robot’s pose at any point in tim=-=e [7, 36]-=-. In combination, however, the problem is much harder. Today, mapping is largely considered the most difficult perceptual problem in robotics. Progress in robot mapping is bound to impact a much broad... |

358 |
Sonar-based real-world mapping and navigation,
- Elfes
- 1987
(Show Context)
Citation Context ... environment, whereas topological maps describe the connectivity of different places. An early representative of the former approach was Elfes and Moravec’s important occupancy grid mapping algorith=-=m [31, 32, 69]-=-, which represents maps by fine-grained grids that model the occupied and free space of the environment. This approach has been used in a great number of robotic systems, such as [8, 9, 10, 42, 83, 98... |

348 | Rao-blackwellised particle filtering for dynamic bayesian networks
- Doucet, Freitas, et al.
- 2000
(Show Context)
Citation Context ...s, such as the FastSLAM algorithm described in [67], promise a reduction to O(log K) complexity for certain situation, by using non-classical statistical sampling techniques for robot path estimation =-=[29, 71]-=- along with efficient tree representations. In practice, the number of features is not known a priori. State-of-the-art implementations often grow this list dynamically. To do so, they maintain a list... |

344 |
Stochastic Models, Estimation, and Control, volume 141-1
- Maybeck
- 1982
(Show Context)
Citation Context ...eflect the actual robot motion. In the field of robot mapping, the single dominating scheme for integrating such temporal data is known as Bayes filter [45], which is highly related to Kalman filters =-=[49, 64]-=-, hidden Markov models [78], dynamic Bayes networks [82], and partially observable Markov decision processes [48, 59, 66, 93]. The Bayes filter extends Bayes rule to temporal estimation problems. It i... |

332 | Incremental mapping of large cyclic environments
- Gutmann, Konolige
- 2000
(Show Context)
Citation Context ... to solve hard correspondence problems. (d) Occupancy grid map built on top of the outcome of the EM mapping algorithm. the correspondence problem. Such large loops are known to be challenging to map =-=[40]-=-. Figure 5c shows the result of applying EM to this data set. The resulting map and path is topologically correct. To illustrate the accuracy of the map, Figures 5b and d show occupancy grid maps buil... |

329 | Experiences with an interactive museum tour-guide robot
- Burgard, Cremers, et al.
- 1999
(Show Context)
Citation Context ...lgorithm [31, 32, 69], which represents maps by fine-grained grids that model the occupied and free space of the environment. This approach has been used in a great number of robotic systems, such as =-=[8, 9, 10, 42, 83, 98, 106, 107]-=-. An alternative metric mapping algorithm was proposed by Chatila and Laumond [15], using sets of polyhedra to describe the geometry of environments. Examples of topological approaches include the wor... |

318 | A real-time algorithm for mobile robot mapping with applications to multi-robot and 3D mapping.
- Thrun, Burgard, et al.
- 2000
(Show Context)
Citation Context ...robot poses are on the left as marked by the letters A, B, and C. mapping, short of the full posterior over maps and poses maintained by Kalman filters. A good example are the algorithms described in =-=[40, 97, 100]-=-. Both of these algorithms use the incremental maximum likelihood approach to build maps, but in addition maintain a posterior distribution over robot poses st. This distribution is calculated using t... |

276 | Collaborative multi-robot exploration.
- Burgard, Fox, et al.
- 2000
(Show Context)
Citation Context ... 21, 43, 88]. Robot exploration in the context of mapping has also been studied extensively. Today’s approaches are usually greedy, that is, they chose control by greedily maximizing information gai=-=n [12, 17, 89, 105]-=-, sometimes under consideration of safety constraints [37]. However, the topic of robot exploration is beyond the scope of this article, hence will not be addressed any further. 3 The Robotic Mapping ... |

257 |
A Survey of Partially Observable Markov Decision Processes: Theory, Models,
- Monahan
- 1982
(Show Context)
Citation Context ...oral data is known as Bayes filter [45], which is highly related to Kalman filters [49, 64], hidden Markov models [78], dynamic Bayes networks [82], and partially observable Markov decision processes =-=[48, 59, 66, 93]-=-. The Bayes filter extends Bayes rule to temporal estimation problems. It is a recursive estimator for computing a sequence of posterior probability distributions over quantities that cannot be observ... |

245 | Optimization of the simultaneous localization and map building algorithm for real time implementation,”
- Guivant, Nebot
- 2001
(Show Context)
Citation Context ...lization and mapping [25, 30], and concurrent mapping and localization [56, 101], respectively. One family of probabilistic approaches employ Kalman filters to estimate the map and the robot location =-=[14, 20, 27, 38, 55, 73, 104]-=-. The resulting maps usually describe the location of landmarks, or significant features in the environment, although recent extensions exist that represent environments by large numbers of raw range ... |

239 | A probabilistic approach to collaborative multi-robot localization. Autonomous robots, 8(3):325– 344,
- Fox, Burgard, et al.
- 2000
(Show Context)
Citation Context ...quired locally by the robots may be unknown. This makes it challenging to estimate the relative location of the robots as they acquire local map information. Moreover, if robots can detect each other =-=[35]-=-, complex correspondence problems may arise related to the identification of individual robots. Another dimension worth exploring is the topic of unstructured environments. Most examples in this surve... |

235 | A probabilistic online mapping algorithm for teams of mobile robots
- Thrun
- 2001
(Show Context)
Citation Context ...andmarks, or significant features in the environment, although recent extensions exist that represent environments by large numbers of raw range measurements [60]. An alternative family of algorithms =-=[23, 86, 87, 97, 101] i-=-s based on Dempster’s expectation maximization algorithm [24, 65]. These approaches specifically address the correspondence problem in mapping, which is the problem of determining whether sensor mea... |

225 | A Survey of Algorithmic Methods for Partially Observed Markov Decision Processes,” - Lovejoy - 1991 |

220 |
Position referencing and consistent world modeling for mobile robots.
- CHATILA, LAUMOND
- 1985
(Show Context)
Citation Context ...he environment. This approach has been used in a great number of robotic systems, such as [8, 9, 10, 42, 83, 98, 106, 107]. An alternative metric mapping algorithm was proposed by Chatila and Laumond =-=[15],-=- using sets of polyhedra to describe the geometry of environments. Examples of topological approaches include the work by Matarić [62], Kuipers [53] and many others [17, 18, 34, 52, 76, 87, 86, 102, ... |

200 | Dynamic map building for an autonomous mobile robot
- Leonard, Durrant-Whyte, et al.
- 1990
(Show Context)
Citation Context ...lization and mapping [25, 30], and concurrent mapping and localization [56, 101], respectively. One family of probabilistic approaches employ Kalman filters to estimate the map and the robot location =-=[14, 20, 27, 38, 55, 73, 104]-=-. The resulting maps usually describe the location of landmarks, or significant features in the environment, although recent extensions exist that represent environments by large numbers of raw range ... |

196 | Probabilistic algorithms and the interactive museum tour-guide robot minerva,”
- Thrun, Beetz, et al.
- 2000
(Show Context)
Citation Context ...lgorithm [31, 32, 69], which represents maps by fine-grained grids that model the occupied and free space of the environment. This approach has been used in a great number of robotic systems, such as =-=[8, 9, 10, 42, 83, 98, 106, 107]-=-. An alternative metric mapping algorithm was proposed by Chatila and Laumond [15], using sets of polyhedra to describe the geometry of environments. Examples of topological approaches include the wor... |

174 |
Topological mapping for mobile robots using a combination of sonar and vision sensing
- Kortenkamp, Weymouth
- 1994
(Show Context)
Citation Context ...oposed by Chatila and Laumond [15], using sets of polyhedra to describe the geometry of environments. Examples of topological approaches include the work by Matarić [62], Kuipers [53] and many others=-= [17, 18, 34, 52, 76, 87, 86, 102, 105, 108]-=-. Topological maps represent environments as a list of significant places that are connected via arcs. Arcs are usually annotated with information on how to navigate from one place to another. However... |

167 | Reinforcement learning in the multi robot domain.
- Mataric
- 1997
(Show Context)
Citation Context ...and are subjects to local maxima. This article placed considerably little emphasis on multi-robot collaboration during mapping. Clearly, many envisioned operational scenarios involved teams of robots =-=[63, 74, 90]-=-, and using multiple collaborative robot for building maps is clearly a worthwhile research goal. Existing techniques work well if the relative starting positions of all robots are known [12, 89], as ... |

167 | Coordination for multi-robot exploration and mapping
- Simmons, Apfelbaum, et al.
- 2000
(Show Context)
Citation Context ... 21, 43, 88]. Robot exploration in the context of mapping has also been studied extensively. Today’s approaches are usually greedy, that is, they chose control by greedily maximizing information gai=-=n [12, 17, 89, 105]-=-, sometimes under consideration of safety constraints [37]. However, the topic of robot exploration is beyond the scope of this article, hence will not be addressed any further. 3 The Robotic Mapping ... |

164 |
Occupancy Grids: A Probabilistic Framework for Robot Perception and Navigation
- Elfes
- 1989
(Show Context)
Citation Context ... environment, whereas topological maps describe the connectivity of different places. An early representative of the former approach was Elfes and Moravec’s important occupancy grid mapping algorith=-=m [31, 32, 69]-=-, which represents maps by fine-grained grids that model the occupied and free space of the environment. This approach has been used in a great number of robotic systems, such as [8, 9, 10, 42, 83, 98... |

161 | A computationally efficient method for large-scale concurrent mapping and localization,”
- Leonard, Feder
- 2000
(Show Context)
Citation Context ...o its growing map. Since then, robotic mapping has commonly been referred to as SLAM or CML, which is short for simultaneous localization and mapping [25, 30], and concurrent mapping and localization =-=[56, 101]-=-, respectively. One family of probabilistic approaches employ Kalman filters to estimate the map and the robot location [14, 20, 27, 38, 55, 73, 104]. The resulting maps usually describe the location ... |

155 | Tracking multiple moving targets with a mobile robot using particle filters and statistical data association.
- Schulz, Burgard, et al.
- 2001
(Show Context)
Citation Context ...ion. The effect is an increase in the location uncertainty of each landmark over time, which subsequently may be counteracted by sensing. Such algorithms are popular in the target tracking literature =-=[4, 68, 85]-=-. Similarly, occupancy grid maps may accommodate certain types of motions, by decaying occupancy over time as discussed in [106]. Extensions of occupancy grids exist that can detect frequently-changin... |

154 | The spmap: A probabilistic framework for simultaneous localization and map building.
- Castellanos, Montiel, et al.
- 1999
(Show Context)
Citation Context ...2], who in 1985 through 1990 proposed a mathematical formulation of the approach that is still in widespread use today. In the following years, a number of researchers developed this approach further =-=[13, 14, 30, 26, 27, 56, 73]-=-, most notably a group of researchers located at the University of Sydney in Australia. In the literature, Kalman filter-based mapping algorithms are often referred to as SLAM algorithms, where SLAM s... |

146 | A distributed model for mobile robot environment-learning and navigation. - Mataric - 1990 |

141 |
Mobile robot localization and map building: A multisensor fusion approach
- Castellanos, Tardos
- 1999
(Show Context)
Citation Context ...lization and mapping [25, 30], and concurrent mapping and localization [56, 101], respectively. One family of probabilistic approaches employ Kalman filters to estimate the map and the robot location =-=[14, 20, 27, 38, 55, 73, 104]-=-. The resulting maps usually describe the location of landmarks, or significant features in the environment, although recent extensions exist that represent environments by large numbers of raw range ... |

136 | Learning topological maps with weak local odometric information,” in
- Shatkay, Kaelbling
- 1997
(Show Context)
Citation Context ...oposed by Chatila and Laumond [15], using sets of polyhedra to describe the geometry of environments. Examples of topological approaches include the work by Matarić [62], Kuipers [53] and many others=-= [17, 18, 34, 52, 76, 87, 86, 102, 105, 108]-=-. Topological maps represent environments as a list of significant places that are connected via arcs. Arcs are usually annotated with information on how to navigate from one place to another. However... |

132 | Conditional particle filters for simultaneous mobile robot localization and people-tracking”,
- Montemerlo, Thrun, et al.
- 2002
(Show Context)
Citation Context ...ion. The effect is an increase in the location uncertainty of each landmark over time, which subsequently may be counteracted by sensing. Such algorithms are popular in the target tracking literature =-=[4, 68, 85]-=-. Similarly, occupancy grid maps may accommodate certain types of motions, by decaying occupancy over time as discussed in [106]. Extensions of occupancy grids exist that can detect frequently-changin... |

119 | The mobile robot Rhino.
- Buhmann, Burgard, et al.
- 1995
(Show Context)
Citation Context |

108 | Map learning and high-speed navigation in rhino.
- Thrun, Bucken, et al.
- 1998
(Show Context)
Citation Context ...ingle sensor measurement zt taken at location st. The literature has devised numerous versions of this probability, for sensors such as sonars, lasers, cameras (mono and stereo), and infrared sensors =-=[72, 99]-=-. Inverse models for range sensors usually attribute high probability of occupancy for grid cells that overlap with detected range, and low probability for grid cells in between this range and the sen... |

98 |
Navigation strategies for exploring indoor environments,”
- Gonzalez-Banos, Latombe
- 2002
(Show Context)
Citation Context ... studied extensively. Today’s approaches are usually greedy, that is, they chose control by greedily maximizing information gain [12, 17, 89, 105], sometimes under consideration of safety constraint=-=s [37]-=-. However, the topic of robot exploration is beyond the scope of this article, hence will not be addressed any further. 3 The Robotic Mapping Problem The problem of robotic mapping is that of acquirin... |

97 | Tracking many objects with many sensors,”
- Pasula, Russell, et al.
- 1999
(Show Context)
Citation Context ...e via maximizing an expected log-likelihood function of the form: θ [i+1] = argmax Ect[log p(m θ t , c t | θ) | θ [i] , m t ] (21) The exact optimization is subject to a mutual exclusivity constra=-=int [23, 75], -=-which specifies that the same object cannot be seen twice in the same map. This constraint is important—as otherwise there may be a tendency to map multiple similar-looking objects to the same objec... |

96 |
Modeling a dynamic environment using a Bayesian multiple hypothesis approach
- Cox, Leonard
- 1994
(Show Context)
Citation Context ...of Gaussian representation [30] or a—highly related—particle filter representation [67, 71]. Such representations have been applied with great success in lower dimensional robot localization probl=-=ems [19, 46, 79, 80]-=-, commonly under the name multi hypothesis Kalman filters. The development of techniques for posterior estimation with unknown correspondence is currently subject of ongoing research, with enormous po... |

94 | Sensor Based Motion Planning: The Hierarchical Generalized Voronoi Graph.
- Choset
- 1996
(Show Context)
Citation Context ...oposed by Chatila and Laumond [15], using sets of polyhedra to describe the geometry of environments. Examples of topological approaches include the work by Matarić [62], Kuipers [53] and many others=-= [17, 18, 34, 52, 76, 87, 86, 102, 105, 108]-=-. Topological maps represent environments as a list of significant places that are connected via arcs. Arcs are usually annotated with information on how to navigate from one place to another. However... |

94 | A solution to the simultaneous localisation and map building (SLAM) problem,”
- Dissanayake, Newman, et al.
- 2001
(Show Context)
Citation Context |

89 | A computationally efficient solution to the simultaneous localisation and map building (SLAM) problem
- Dissanayake, Durrant-Whyte, et al.
- 2000
(Show Context)
Citation Context ...induced problem of localizing the robot relative to its growing map. Since then, robotic mapping has commonly been referred to as SLAM or CML, which is short for simultaneous localization and mapping =-=[25, 30]-=-, and concurrent mapping and localization [56, 101], respectively. One family of probabilistic approaches employ Kalman filters to estimate the map and the robot location [14, 20, 27, 38, 55, 73, 104]... |

89 |
Rao-Blackwellised particle filtering for dynamic Bayesian networks.
- Murphy, Russell
- 2001
(Show Context)
Citation Context ...s, such as the FastSLAM algorithm described in [67], promise a reduction to O(log K) complexity for certain situation, by using non-classical statistical sampling techniques for robot path estimation =-=[29, 71]-=- along with efficient tree representations. In practice, the number of features is not known a priori. State-of-the-art implementations often grow this list dynamically. To do so, they maintain a list... |

88 |
Navigating Mobile Robots: Systems
- Borenstein, Everett, et al.
- 1996
(Show Context)
Citation Context ...y further. 3 The Robotic Mapping Problem The problem of robotic mapping is that of acquiring a spatial model of a robot’s environment. Maps are commonly used for robot navigation (e.g., localization=-=) [7, 51]-=-. To acquire a map, robots must possess sensors that enable it to perceive the outside world. Sensors commonly brought to bear for this task include cameras, range finders using sonar, laser, and infr... |

85 |
Stochastic Processes and Filtering Theory,
- Jazwinsky
- 1970
(Show Context)
Citation Context ...ad of controls, since they may more accurately reflect the actual robot motion. In the field of robot mapping, the single dominating scheme for integrating such temporal data is known as Bayes filter =-=[45]-=-, which is highly related to Kalman filters [49, 64], hidden Markov models [78], dynamic Bayes networks [82], and partially observable Markov decision processes [48, 59, 66, 93]. The Bayes filter exte... |

85 | Interactive construction of 3D models from panoramic mosaics
- Shum, Han, et al.
- 1998
(Show Context)
Citation Context ...], of furniture and other objects that move [6, 83]. Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection that is still somewhat underexploite=-=d [1, 3, 5, 16, 21, 43, 88]. -=-Robot exploration in the context of mapping has also been studied extensively. Today’s approaches are usually greedy, that is, they chose control by greedily maximizing information gain [12, 17, 89,... |

81 | Towards personal service robots for the elderly.
- Baltus, Fox, et al.
- 2000
(Show Context)
Citation Context ...ical framework accompanying such methods. Clearly, the issue has to be addressed if robots shall operate in environments over long periods of time, such as the long-envisioned personal service robots =-=[81, 33, 54, 84]-=-. A viable research goal, thus, would be to develop a lifelong mapping robot, one that is capable of continually 23supdating its map (and its object models) over its lifetime. We also need technologie... |

68 |
Error correction in mobile robot map learning,”
- Engelson, McDermott
- 1992
(Show Context)
Citation Context |

67 |
The Structure and Solution of the Simultaneous Localisation and Map Building Problem,”
- Newman
- 1999
(Show Context)
Citation Context |

64 | Exploration and model building in mobile robot domains.
- Thrun
- 1993
(Show Context)
Citation Context ...icient maximum likelihood estimates. One of the most common approaches, which from a mathematical point of view is inferior to both Kalman filters and EM, is the incremental maximum likelihood method =-=[32, 69, 94, 105, 107]-=-. The basic idea is to incrementally build a single map as the sensor data arrives, but without keeping track of any residual uncertainty. Such a methodology can be viewed as a M-step in EM, without a... |

63 | Semiautomatic 3-D model extraction from uncalibrated 2-D camera views.
- Becker, Bove
- 1995
(Show Context)
Citation Context ...], of furniture and other objects that move [6, 83]. Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection that is still somewhat underexploite=-=d [1, 3, 5, 16, 21, 43, 88]. -=-Robot exploration in the context of mapping has also been studied extensively. Today’s approaches are usually greedy, that is, they chose control by greedily maximizing information gain [12, 17, 89,... |

62 | On the design of behavior-based multi-robot teams.
- Parker
- 1996
(Show Context)
Citation Context ...and are subjects to local maxima. This article placed considerably little emphasis on multi-robot collaboration during mapping. Clearly, many envisioned operational scenarios involved teams of robots =-=[63, 74, 90]-=-, and using multiple collaborative robot for building maps is clearly a worthwhile research goal. Existing techniques work well if the relative starting positions of all robots are known [12, 89], as ... |

59 |
editors, AI-based Mobile Robots: Case studies of successful robot systems,
- Kortenkamp, Bonasso, et al.
- 1998
(Show Context)
Citation Context ...y further. 3 The Robotic Mapping Problem The problem of robotic mapping is that of acquiring a spatial model of a robot’s environment. Maps are commonly used for robot navigation (e.g., localization=-=) [7, 51]-=-. To acquire a map, robots must possess sensors that enable it to perceive the outside world. Sensors commonly brought to bear for this task include cameras, range finders using sonar, laser, and infr... |

49 | Bayesian estimation and Kalman filtering: A unified framework for mobile robot localization,
- Roumeliotis, Bekey
- 2000
(Show Context)
Citation Context ...of Gaussian representation [30] or a—highly related—particle filter representation [67, 71]. Such representations have been applied with great success in lower dimensional robot localization probl=-=ems [19, 46, 79, 80]-=-, commonly under the name multi hypothesis Kalman filters. The development of techniques for posterior estimation with unknown correspondence is currently subject of ongoing research, with enormous po... |

48 | Natural Communication with Robots”
- Torrance
- 1994
(Show Context)
Citation Context |

46 | Active Global Localisation for a Mobile Robot Using Multiple Hypothesis Tracking.
- Jensfelt, Kristensen
- 1999
(Show Context)
Citation Context ...of Gaussian representation [30] or a—highly related—particle filter representation [67, 71]. Such representations have been applied with great success in lower dimensional robot localization probl=-=ems [19, 46, 79, 80]-=-, commonly under the name multi hypothesis Kalman filters. The development of techniques for posterior estimation with unknown correspondence is currently subject of ongoing research, with enormous po... |

46 |
Using EM to learn 3D models with mobile robots
- Liu, Emery, et al.
- 2001
(Show Context)
Citation Context ...oints in time correspond to the same physical entity in the real world. A third family of probabilistic techniques seek to identify objects in the environment, which may correspond to ceilings, walls =-=[44, 58, 61], -=-doors that might be open or closed [2], of furniture and other objects that move [6, 83]. Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection... |

44 |
An experimental and theoretical investigation into simultaneous localisation and map building
- Dissanayake, Newman, et al.
- 1999
(Show Context)
Citation Context ...2], who in 1985 through 1990 proposed a mathematical formulation of the approach that is still in widespread use today. In the following years, a number of researchers developed this approach further =-=[13, 14, 30, 26, 27, 56, 73]-=-, most notably a group of researchers located at the University of Sydney in Australia. In the literature, Kalman filter-based mapping algorithms are often referred to as SLAM algorithms, where SLAM s... |

42 |
A Bayesian algorithm for simultaneous localization and map building. Unpublished manuscript
- Durrant-Whyte, Majumder, et al.
- 2001
(Show Context)
Citation Context ...induced problem of localizing the robot relative to its growing map. Since then, robotic mapping has commonly been referred to as SLAM or CML, which is short for simultaneous localization and mapping =-=[25, 30]-=-, and concurrent mapping and localization [56, 101], respectively. One family of probabilistic approaches employ Kalman filters to estimate the map and the robot location [14, 20, 27, 38, 55, 73, 104]... |

41 |
The Application of Robotics to A Mobility Aid for The Elderly Blind,”
- Lacey, Dawson-Howe
- 1998
(Show Context)
Citation Context ...ical framework accompanying such methods. Clearly, the issue has to be addressed if robots shall operate in environments over long periods of time, such as the long-envisioned personal service robots =-=[81, 33, 54, 84]-=-. A viable research goal, thus, would be to develop a lifelong mapping robot, one that is capable of continually 23supdating its map (and its object models) over its lifetime. We also need technologie... |

37 | 2000. Integration of range and image sensing for photorealistic 3D modeling
- Allen, Stamos
(Show Context)
Citation Context |

37 | Many robots make short work
- Guzzoni, Cheyer, et al.
- 1997
(Show Context)
Citation Context |

35 | Learning to explore and build maps.
- Pierce, Kuipers
- 1994
(Show Context)
Citation Context |

34 | Visually realistic mapping of a planar environment with stereo
- Iocchi, Konolige, et al.
- 2000
(Show Context)
Citation Context ...oints in time correspond to the same physical entity in the real world. A third family of probabilistic techniques seek to identify objects in the environment, which may correspond to ceilings, walls =-=[44, 58, 61], -=-doors that might be open or closed [2], of furniture and other objects that move [6, 83]. Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection... |

32 |
Sonar-based mapping of large-scale mobile robot environments using EM
- Burgard, Fox, et al.
- 1999
(Show Context)
Citation Context ...algorithm [65]. Applied to the robotic mapping problem, the EM algorithm has quite orthogonal characteristics. EM algorithms constitute today’s best solutions to the correspondence problem in mappin=-=g [11, 101]-=-. In particular, EM algorithms have been found to generate consistent maps of large-scale cyclic environment even if all features look alike and cannot be distinguished perceptually. However, EM algor... |

30 | Coordinated deployment of multiple, heterogeneous robots
- Simmons, Apfelbaum, et al.
- 2000
(Show Context)
Citation Context ...and are subjects to local maxima. This article placed considerably little emphasis on multi-robot collaboration during mapping. Clearly, many envisioned operational scenarios involved teams of robots =-=[63, 74, 90]-=-, and using multiple collaborative robot for building maps is clearly a worthwhile research goal. Existing techniques work well if the relative starting positions of all robots are known [12, 89], as ... |

29 |
chain flipping for structure from motion with unknown correspondence
- “EM
- 2000
(Show Context)
Citation Context ...andmarks, or significant features in the environment, although recent extensions exist that represent environments by large numbers of raw range measurements [60]. An alternative family of algorithms =-=[23, 86, 87, 97, 101] i-=-s based on Dempster’s expectation maximization algorithm [24, 65]. These approaches specifically address the correspondence problem in mapping, which is the problem of determining whether sensor mea... |

28 | Simultaneous Localisation and Map Building
- Csorba
- 1997
(Show Context)
Citation Context |

26 | Learning models for robot navigation
- Shatkay
- 1998
(Show Context)
Citation Context |

24 |
Navigation mobiler roboter mit laserscans
- Gutmann, Nebel
- 1997
(Show Context)
Citation Context ...ally lack detailed geometric descriptions of the environment. A recent extension of the basic paradigm is known as the Lu/Milios algorithm [60]. This algorithm was successfully implemented by Gutmann =-=[39, 41]-=-. The Lu/Milios algorithm is somewhat specific to laser range data. It combines two basic estimation phases: A phase where Kalman filters are used to calculate posteriors over maps, and another where ... |

22 | Learning occupancy grids with forward models
- Thrun
(Show Context)
Citation Context ...unless the robot is in motion. A final, more subtle deficiency stems from the independence assumption between multiple grid cells. Such an assumption, while convenient, can lead to inferior maps. See =-=[96]-=- for more details. 9 Object Maps Another family of mapping algorithms described in this paper addresses the problem of building maps composed of basic geometric shapes or objects, such as lines, walls... |

21 |
Towards object mapping in dynamic environments with mobile robots
- Biswas, Limketkai, et al.
- 2002
(Show Context)
Citation Context ...bilistic techniques seek to identify objects in the environment, which may correspond to ceilings, walls [44, 58, 61], doors that might be open or closed [2], of furniture and other objects that move =-=[6, 83]. -=-Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection that is still somewhat underexploited [1, 3, 5, 16, 21, 43, 88]. Robot exploration in the... |

19 |
Online acquisition of compact volumetric maps with mobile robots
- Martin, Thrun
- 2002
(Show Context)
Citation Context ...oints in time correspond to the same physical entity in the real world. A third family of probabilistic techniques seek to identify objects in the environment, which may correspond to ceilings, walls =-=[44, 58, 61], -=-doors that might be open or closed [2], of furniture and other objects that move [6, 83]. Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection... |

18 |
Robot navigation by 3D spatial evidence grids.
- Moravec, Martin
- 1994
(Show Context)
Citation Context ...n. The surface structure in Figure 10 is extremely rugged, reflecting the noise in the sensor measurements. One might be tempted to apply occupancy grid-style techniques for reducing the noise, as in =-=[70]-=-. However, in this specific data set each feature in the environment is sensed at most once, whereas techniques like occupancy grids require that a grid cell can be measured many times, so that inform... |

16 | A probabilistic technique for simultaneous localization and door state estimation with mobile robots in dynamic environments.
- Avots, Lim, et al.
- 2002
(Show Context)
Citation Context ...ity in the real world. A third family of probabilistic techniques seek to identify objects in the environment, which may correspond to ceilings, walls [44, 58, 61], doors that might be open or closed =-=[2], -=-of furniture and other objects that move [6, 83]. Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection that is still somewhat underexploited [... |

16 |
3D reconstruction of environments for virtual reconstruction
- Bajcsy, Kamberova, et al.
- 2000
(Show Context)
Citation Context ...], of furniture and other objects that move [6, 83]. Many of these technique have counterparts in the computer vision and photogrammetry literature—-a connection that is still somewhat underexploited =-=[1, 3, 5, 16, 21, 43, 88]-=-. Robot exploration in the context of mapping has also been studied extensively. Today’s approaches are usually greedy, that is, they chose control by greedily maximizing information gain [12, 17, 89,... |

10 |
Interpreting stereo vision for a mobile robot
- Murray, Little
- 1998
(Show Context)
Citation Context ...ingle sensor measurement zt taken at location st. The literature has devised numerous versions of this probability, for sensors such as sonars, lasers, cameras (mono and stereo), and infrared sensors =-=[72, 99]-=-. Inverse models for range sensors usually attribute high probability of occupancy for grid cells that overlap with detected range, and low probability for grid cells in between this range and the sen... |

9 |
A mobile robot sense net
- Konolige, Gutmann, et al.
- 1999
(Show Context)
Citation Context ...ementations appear to work well in real-time when used in office-building type environments. The hybrid mapping algorithm has been extended to handle multiple robots that jointly acquire a single map =-=[50, 97]-=-. Figure 8 plots a map acquired by three autonomous robots, which coordinated their explo16s(a) (b) Figure 9: (a) Occupancy grid map and (b) architectural blueprint of a recently constructed building.... |

9 |
Mobile robot self-localization using PDAB
- Reuter
(Show Context)
Citation Context |

8 | Is robotics going statistics? The field of probabilistic robotics
- Thrun
- 2001
(Show Context)
Citation Context ...on. This development parallels a much broader trend in mobile robotics, where probabilistic techniques are commonly the method of choice over more ad hoc approaches, such as behavior-based techniques =-=[95]-=-. Overall, the situation in robot mapping is encouraging. After approximately two decades of research, the field has matured to a point where detailed maps of complex environments can be built in real... |

7 |
Sensorinterpretation und Kartenerstellung für mobile Roboter
- Schneider
- 1994
(Show Context)
Citation Context |

6 |
Sensor Based Planning: The Hierarhical Generalized Voronoi Graph
- Choset, Burdick
- 1996
(Show Context)
Citation Context |

4 | Three-dimensional reconstruction of points and lines with unknown correspondence across images
- Cheng, Riseman, et al.
- 2000
(Show Context)
Citation Context |

4 |
Sensor based creation of indoor virtual environment models
- Hakim, Boulanger
- 1997
(Show Context)
Citation Context |

1 |
Robuste Navigation Autonomer Mobiler Systeme. Akademische Verlagsgesellschaft Aka
- Gutmann
- 2000
(Show Context)
Citation Context ...ally lack detailed geometric descriptions of the environment. A recent extension of the basic paradigm is known as the Lu/Milios algorithm [60]. This algorithm was successfully implemented by Gutmann =-=[39, 41]-=-. The Lu/Milios algorithm is somewhat specific to laser range data. It combines two basic estimation phases: A phase where Kalman filters are used to calculate posteriors over maps, and another where ... |