Person:
Dellaert, Frank

Permanent Link
https://hdl.handle.net/1853/71202
Associated Organization(s)
Organizational Unit
ORCID
ArchiveSpace Name Record

Filters

2 1
2009 2
2
1 1 1
2
Reset filters

Settings
Author: Kaess, Michael × End date: 2009 × Start date: 2009 × Has files: Yes ×

Publication Search Results

Now showing 1 - 2 of 2
Thumbnail Image
Item

Flow Separation for Fast and Robust Stereo Odometry

2009-05 , Kaess, Michael , Ni, Kai , Dellaert, Frank

Separating sparse flow provides fast and robust stereo visual odometry that deals with nearly degenerate situations that often arise in practical applications.We make use of the fact that in outdoor situations different constraints are provided by close and far structure, where the notion of close depends on the vehicle speed. The motion of distant features determines the rotational component that we recover with a robust two-point algorithm. Once the rotation is known, we recover the translational component from close features using a robust one-point algorithm. The overall algorithm is faster than estimating the motion in one step by a standard RANSAC-based three-point algorithm. And in contrast to other visual odometry work, we avoid the problem of nearly degenerate data, under which RANSAC is known to return inconsistent results. We confirm our claims on data from an outdoor robot equipped with a stereo rig.

View
Thumbnail Image
Item

Covariance Recovery from a Square Root Information Matrix for Data Association

2009 , Kaess, Michael , Dellaert, Frank

Data association is one of the core problems of simultaneous localization and mapping (SLAM), and it requires knowledge about the uncertainties of the estimation problem in the form of marginal covariances. However, it is often difficult to access these quantities without calculating the full and dense covariance matrix, which is prohibitively expensive. We present a dynamic programming algorithm for efficient recovery of the marginal covariances needed for data association. As input we use a square root information matrix as maintained by our incremental smoothing and mapping (iSAM) algorithm. The contributions beyond our previous work are an improved algorithm for recovering the marginal covariances and a more thorough treatment of data association now including the joint compatibility branch and bound (JCBB) algorithm. We further show how to make information theoretic decisions about measurements before actually taking the measurement, therefore allowing a reduction in estimation complexity by omitting uninformative measurements. We evaluate our work on simulated and real-world data.

View