Towards Lazy Data Association in SLAM

Hähnel, Dirk; Thrun, Sebastian; Wegbreit, Ben; Burgard, Wolfram

doi:10.1007/11008941_45

Cited by 64 publications

(62 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The latter has been shown by several other researchers (e.g. [19]), and the former can be accomplished by simply re-running the present algorithm as new loops in the pose network are closed.…”

Section: Discussionmentioning

confidence: 56%

Efficient Construction of Globally Consistent Ladar Maps Using Pose Network Topology and Nonlinear Programming

Kelly

Unnikrishnan

2005

Springer Tracts in Advanced Robotics

View full text Add to dashboard Cite

Abstract. Many instances of the mobile robot guidance mapping problem exhibit a topology that can be represented as a graph of nodes (observations) connected by edges (poses). We show that a cycle basis of this pose network can be used to generate the independent constraint equations in a natural constrained optimization formulation of the mapping problem. Explicit reasoning about the loop topology of the network can automatically generate such a cycle basis in linear time. Furthermore, in many practical cases, the pose network has sparse structure and the associated equations can then be solved time linear in the number of images. This approach can be used to construct globally consistent maps on very large scales in very limited computation. While the technique is applicable to mapbuilding in general, and even optimization in general, it is illustrated here for batch processing of 2D ladar scans into a mobile robot guidance map.

show abstract

Section: Discussionmentioning

confidence: 56%

Efficient Construction of Globally Consistent Ladar Maps Using Pose Network Topology and Nonlinear Programming

Kelly

Unnikrishnan

2005

Springer Tracts in Advanced Robotics

View full text Add to dashboard Cite

show abstract

“…The relative error between the spline generated by Algorithm D.4.1 at iteration k and the true curve is defined as e k = r k −r true / r true (D. 33) wherer k is the discretized smoothing spline output, andr true is the corresponding true curve in polar coordinates.…”

Section: D54 Resultsmentioning

confidence: 99%

“…The magnitude of N is set in the prefiltering step, and while computation time benefits from a small value, a larger one may yield a more rapid convergence, allowing for fewer iterations. In SLAM applications, large association matrices are often inverted online, see for instance [33]. For online applications of Algorithm D.4.1 a similar approach should be used for inversion of the matrix H. As the aim of the simulations and experiments in this work was to evaluate accuracy and convergence properties of Algorithm D.4.1, time-optimization of the code remains for future work.…”

Section: D53 Computation Timementioning

confidence: 99%

Contour reconstruction using recursive smoothing splines - Algorithms and experimental validation

Karasalo

Piccolo

Kragić

et al. 2009

Robotics and Autonomous Systems

View full text Add to dashboard Cite

In this paper, a recursive smoothing spline approach for contour reconstruction is studied and evaluated. Periodic smoothing splines are used by a robot to approximate the contour of encountered obstacles in the environment. The splines are generated through minimizing a cost function subject to constraints imposed by a linear control system and accuracy is improved iteratively using a recursive spline algorithm. The filtering effect of the smoothing splines allows for usage of noisy sensor data and the method is robust with respect to odometry drift. The algorithm is extensively evaluated in simulations for various contours and in experiments using a SICK laser scanner mounted on a PowerBot from ActivMedia Robotics.Keywords: mapping, optimal control, recursive smoothing splines, implementation D.1 IntroductionToday robots are expected to operate in at least partially unknown, open-ended environments, detecting and moving around obstacles, building maps of the environment and localizing their position. The area of autonomous mapping has during the past few years matured and there has been a significant amount of work reported using both visual and laser sensors, [8], [26]. However, most methods are still concerned with detecting and maintaining a dense set of discrete features of the environment that are suitable for navigation and localization without retrieving any specific knowledge about the detailed shape of the encountered objects.In this paper, we consider the problem of estimating representations of objects using a type of continuous closed curves, the recursive periodic smoothing splines. The splines are retrieved from noisy measurements of the true contour. Intended applications include mapping, identification and path planning. It is well known that interpolating splines from noisy measurement data will give a poor result, as the resulting curve will go through every measurement point. Data smoothing has been a classical problem in system and control history [1], [13], [18]. The theory of regular smoothing splines is thoroughly treated in [29] and [30]. It has been further shown in [35] that control theoretic smoothing splines, where the curve is found through minimizing a cost function, act as filters and are better suited for noisy measurements. It is also noted in [27] that smoothing 89

show abstract

“…This turns − log p(X ) into a quadratic function over the variables X . Setting the first derivative to zero yields the desired minimum in closed form, as described in more detail in [17] for details. We also note that there are a number of alternative techniques for minimizing − log p(X ), some of which exploit the sparse nature of the potentials [15,30,45] …”

Section: Simultaneous Localization and Mappingmentioning

confidence: 99%

“…Figure 8c illustrates the potential outcome of this approach: in this example, a different sequence of data association decisions yields a better map. As described in detail in [17], adding and removing consistency constraints can be done efficiently, and calculating the resulting configuration X does not require a full solution of the optimization problem.…”

Section: Data Associationmentioning

confidence: 99%

Autonomous exploration and mapping of abandoned mines

Thrun

Thayer²,

Whittaker³

et al. 2004

IEEE Robot. Automat. Mag.

188

View full text Add to dashboard Cite

Abandoned mines pose significant threats to society, yet a large fraction of them lack accurate maps. This article discusses the software architecture of an autonomous robotic system designed to explore and map abandoned mines. We have built a robot capable of autonomously exploring abandoned mines. A new set of software tools is presented, enabling robots to acquire maps of unprecedented size and accuracy. On May 30, 2003, our robot "Groundhog" successfully explored and mapped a main corridor of the abandoned Mathies mine near Courtney, PA. The article also discusses some of the challenges that arise in the subterraneans environments, and some the difficulties of building truly autonomous robots.

show abstract

Towards Lazy Data Association in SLAM

Cited by 64 publications

References 22 publications

Efficient Construction of Globally Consistent Ladar Maps Using Pose Network Topology and Nonlinear Programming

Efficient Construction of Globally Consistent Ladar Maps Using Pose Network Topology and Nonlinear Programming

Contour reconstruction using recursive smoothing splines - Algorithms and experimental validation

Autonomous exploration and mapping of abandoned mines

Contact Info

Product

Resources

About