KALMANSAC: robust filtering by consensus

Vedaldi, Andrea; Jin, Hailin; Favaro, Paolo; Soatto, Stefano

doi:10.1109/iccv.2005.130

Cited by 30 publications

(21 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a consequence of this, the number of random hypothesis necessary to obtain a mismatch-free subset is reduced by several orders of magnitude. Incorporating probabilistic predictions into RANSAC has been previously investigated in [15] for the case of weak priors and also in an EKF context [28]. Nevertheless, the reduction of the subset size and hence full exploitation of the strong a priori information available in the EKF is not explored in these works.…”

mentioning

confidence: 99%

1-point RANSAC for EKF-based Structure from Motion

Civera

Grasa

Davison

et al. 2009

2009 IEEE/RSJ International Conference on Intelligent Robots and Systems

102

View full text Add to dashboard Cite

Abstract-Recently, classical pairwise Structure From Motion (SfM) techniques have been combined with non-linear global optimization (Bundle Adjustment, BA) over a sliding window to recursively provide camera pose and feature location estimation from long image sequences. Normally called Visual Odometry, these algorithms are nowadays able to estimate with impressive accuracy trajectories of hundreds of meters; either from an image sequence (usually stereo) as the only input, or combining visual and propioceptive information from inertial sensors or wheel odometry.This paper has a double objective. First, we aim to illustrate for the first time how similar accuracy and trajectory length can be achieved by filtering-based visual SLAM methods. Specifically, a camera-centered Extended Kalman Filter is used here to process a monocular sequence as the only input, with 6DOF motion estimated. Features are kept live in the filter while visible as the camera explores forward, and are deleted from the state once they go out of view.This permits an increase in the number of tracked features per frame from tens to around a hundred. While improving the accuracy of the estimation, it makes computationally infeasible the exhaustive Branch and Bound search performed by standard JCBB for match outlier rejection. As a second contribution that overcomes this problem, we present here a RANSAC-like algorithm that exploits the probabilistic prediction of the filter. This use of prior information makes it possible to reduce the size of the minimal data subset to instantiate a hypothesis to the minimum possible of 1 point, greatly increasing the efficiency of the outlier rejection stage.Experimental results from real image sequences covering trajectories of hundreds of meters are presented and compared against RTK GPS ground truth. Estimation errors are about 1% of the trajectory for trajectories up to 650 metres.

show abstract

mentioning

confidence: 99%

1-point RANSAC for EKF-based Structure from Motion

Civera

Grasa

Davison

et al. 2009

2009 IEEE/RSJ International Conference on Intelligent Robots and Systems

102

View full text Add to dashboard Cite

show abstract

“…In other words, we found the SRD-SPKF to be more sensitive to outliers than the S-SPKF. Although RANSAC could be applied to the SRD-SPKF, as it has been for the EKF [23], we note that our S-SPKF would only help RANSAC, as it is less sensitive to outliers. It is not our contention that re-linearizing will always be problematic or less accurate.…”

Section: Discussionmentioning

confidence: 99%

A Serial Approach to Handling High-Dimensional Measurements in the Sigma-Point Kalman Filter

McManus¹,

Barfoot²

2011

Robotics: Science and Systems VII

View full text Add to dashboard Cite

Abstract-Pose estimation is a critical skill in mobile robotics and is often accomplished using onboard sensors and a Kalman filter estimation technique. For systems to run online, computational efficiency of the filter design is crucial, especially when faced with limited computing resources. In this paper, we present a novel approach to serially process high-dimensional measurements in the Sigma-Point Kalman Filter (SPKF), in order to achieve a low computational cost that is linear is the measurement dimension. Although the concept of serially processing measurements has been around for quite some time in the context of the Extended Kalman Filter (EKF), few have considered this approach with the SPKF. At first glance, it may be tempting to apply the SPKF update step serially. However, we prove that without re-drawing sigma points, this 'naive' approach cannot guarantee the positive-definiteness of the state covariance matrix (not the case for the EKF). We then introduce a novel method for the Sigma-Point Kalman Filter to process high-dimensional, uncorrelated measurements serially that is algebraically equivalent to processing the measurements in parallel, but still achieves a computational cost linear in the measurement dimension.

show abstract

“…A rather similar approach is the KALMANSAC ( [20]), where an explicit estimation of the process related to the outliers is added. Nevertheless, in the absolute orientation problem we are not interested in an explicit estimation of the range data but in the localization (or "relocation") of the LIDAR.…”

Section: Morp-a064mentioning

confidence: 99%

A Fast RANSAC-Based Registration Algorithm for Accurate Localization in Unknown Environments using LIDAR Measurements

Fontanelli

Ricciato

Soatto

2007

2007 IEEE International Conference on Automation Science and Engineering

View full text Add to dashboard Cite

Abstract-The problem of accurate localization using only measurements from a LIDAR sensor is analyzed in this paper. The sensor is rigidly fixed on a generic moving platform, which moves on a plane. Practical on-line applications of localization algorithms impose constraints on the execution time, problem that is addressed in this paper and compared with other existing solutions.Due to the nature of the sensor adopted, the localization algorithm is based on a fast and accurate registration algorithm, which is able to deal with noisy measurements, outliers and dynamic environments. The proposed solution relies on the RANSAC algorithm in combination with a Huber kernel in order to cope with typical nuisances in LIDAR measurements. The robust registration is successively used in combination with an Extended Kalman Filter to track the trajectory of the LIDAR over time, hence to solve the localization problem.Simulations and experimental results are reported to show the feasibility of the proposed approach.

show abstract

KALMANSAC: robust filtering by consensus

Cited by 30 publications

References 18 publications

1-point RANSAC for EKF-based Structure from Motion

1-point RANSAC for EKF-based Structure from Motion

A Serial Approach to Handling High-Dimensional Measurements in the Sigma-Point Kalman Filter

A Fast RANSAC-Based Registration Algorithm for Accurate Localization in Unknown Environments using LIDAR Measurements

Contact Info

Product

Resources

About