FrameSLAM: From Bundle Adjustment to Real-Time Visual Mapping

Konolige, Kurt; Agrawal, Motilal

doi:10.1109/tro.2008.2004832

Cited by 451 publications

(305 citation statements)

References 36 publications

Supporting

Mentioning

303

Contrasting

Unclassified

Order By: Relevance

“…Probably the most complete is that of Mei et al [18], which combines robust and accurate local visual odometry with constant-time large-scale mapping (enabled by a continuous relative map representation), highperformance appearance-based loop closure detection, and global metric map optimisation if required. Another similar system is that presented by Konolige and Agrawal [16].…”

Section: Introductionmentioning

confidence: 93%

Scale Drift-Aware Large Scale Monocular SLAM

Strasdat¹,

Montiel²,

Davison³

2010

Robotics: Science and Systems VI

358

309

View full text Add to dashboard Cite

Abstract-State of the art visual SLAM systems have recently been presented which are capable of accurate, large-scale and real-time performance, but most of these require stereo vision. Important application areas in robotics and beyond open up if similar performance can be demonstrated using monocular vision, since a single camera will always be cheaper, more compact and easier to calibrate than a multi-camera rig.With high quality estimation, a single camera moving through a static scene of course effectively provides its own stereo geometry via frames distributed over time. However, a classic issue with monocular visual SLAM is that due to the purely projective nature of a single camera, motion estimates and map structure can only be recovered up to scale. Without the known inter-camera distance of a stereo rig to serve as an anchor, the scale of locally constructed map portions and the corresponding motion estimates is therefore liable to drift over time.In this paper we describe a new near real-time visual SLAM system which adopts the continuous keyframe optimisation approach of the best current stereo systems, but accounts for the additional challenges presented by monocular input. In particular, we present a new pose-graph optimisation technique which allows for the efficient correction of rotation, translation and scale drift at loop closures. Especially, we describe the Lie group of similarity transformations and its relation to the corresponding Lie algebra. We also present in detail the system's new image processing front-end which is able accurately to track hundreds of features per frame, and a filter-based approach for feature initialisation within keyframe-based SLAM. Our approach is proven via large-scale simulation and real-world experiments where a camera completes large looped trajectories.

show abstract

Section: Introductionmentioning

confidence: 93%

Scale Drift-Aware Large Scale Monocular SLAM

Strasdat¹,

Montiel²,

Davison³

2010

Robotics: Science and Systems VI

358

309

View full text Add to dashboard Cite

show abstract

“…The authors conclude (with some reservations), that key frame optimization is superior compared to filtering in terms of accuracy per computing time. Methods that combine both approaches in order to take advantage of their individual strengths are iSAM [KRD08], FrameSLAM [KA08], and the stereo SLAM system of [MSC + 09].…”

Section: Keyframe Slammentioning

confidence: 99%

Mobile Visual Location Recognition

Schroth¹,

Huitl

Chen

et al. 2011

IEEE Signal Process. Mag.

107

View full text Add to dashboard Cite

“…Visual SLAM is attractive because it uses the available on-board cameras to complete the SLAM task. The current successful VSLAM implementations use Extended Kalman Filter, (EKF) [5], [6] or Bundle Adjustment, (BA) [7], [8] to solve the problem. There are many differences in computations side, but on the accuracy side, filtering techniques may reach comparable accuracy as that of BA [9].…”

Section: Introductionmentioning

confidence: 99%

Improving visual SLAM accuracy through deliberate camera oscillations

Heshmat

Abdellatif

Abbas

2013

2013 IEEE International Symposium on Robotic and Sensors Environments (ROSE)

View full text Add to dashboard Cite

Abstract-Visual Simultaneous Localization And Mapping, (VSLAM) algorithms exploit the observation of scene naturallyexisting distinct features to infer the camera motion and build a map of a static environment. There is an increasing interest towards building efficient VSLAM algorithms mainly from computational perspectives; however, there may be insufficient clues to solve for SLAM parameters efficiently. In this paper, deliberate camera oscillations are superimposed on the camera main motion (robot motion), mostly in a lateral direction to give sufficient physical clues for the solution. Filtering methods exploit correlation to infer the motion parameters, and since oscillation introduces more local changes, it can enhance the estimation by correlation. Simulation results are presented showing the effects of oscillation parameters on visual SLAM performance in different motion scenarios. The results showed significant improvement of accuracy for oscillating camera over the steady camera case, and in several cases errors are reduced to less than half its value. These simulation results can be the basis to design a real experimental system.

show abstract

FrameSLAM: From Bundle Adjustment to Real-Time Visual Mapping

Cited by 451 publications

References 36 publications

Scale Drift-Aware Large Scale Monocular SLAM

Scale Drift-Aware Large Scale Monocular SLAM

Mobile Visual Location Recognition

Improving visual SLAM accuracy through deliberate camera oscillations

Contact Info

Product

Resources

About