2018 IEEE/OES Autonomous Underwater Vehicle Workshop (AUV) 2018
DOI: 10.1109/auv.2018.8729731
|View full text |Cite
|
Sign up to set email alerts
|

A Comparison of Submap Registration Methods for Multibeam Bathymetric Mapping

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
5
0

Year Published

2019
2019
2024
2024

Publication Types

Select...
4
3

Relationship

3
4

Authors

Journals

citations
Cited by 7 publications
(5 citation statements)
references
References 23 publications
0
5
0
Order By: Relevance
“…A version of the generalized iterative closest point (GICP) from [19] restricted to x,y and yaw is applied to find the relative transformation T GICP ∈ R 3 between the submap frames that minimizes the plane-to-plane distance between the point clouds. GICP is more suitable for the registration of large, mostly flat submaps [20] than the variants of ICP typically used on other more structured underwater environments ( [11], [18], [17]). This is because GICP models locally the surface from both point clouds during the matching step, easing the registration of large planar surfaces with minimal overlap and scarce features.…”
Section: Slam With Bathymetric Submapsmentioning
confidence: 99%
“…A version of the generalized iterative closest point (GICP) from [19] restricted to x,y and yaw is applied to find the relative transformation T GICP ∈ R 3 between the submap frames that minimizes the plane-to-plane distance between the point clouds. GICP is more suitable for the registration of large, mostly flat submaps [20] than the variants of ICP typically used on other more structured underwater environments ( [11], [18], [17]). This is because GICP models locally the surface from both point clouds during the matching step, easing the registration of large planar surfaces with minimal overlap and scarce features.…”
Section: Slam With Bathymetric Submapsmentioning
confidence: 99%
“…In the remainder of this paper we apply the presented general approach for learning Gaussian distributions from unordered sets of points to the specific problem of bathymetric graph SLAM with GICP registration as introduced in [21]. The reason to focus on GICP as opposed to ICP is that this method works better on the kind of bathymetric point clouds produced from surveys of unstructured seabed, as discussed in [6].…”
Section: Methodsmentioning
confidence: 99%
“…Of the various types of sonar, multibeam echo sounders (MBES) provide the most suitable type of raw data for SLAM methods. This data is essentially a point cloud sampled from the bathymetric surface, and applying registration methods to these measurements is a well-studied problem [6]. However, when fusing the output of the registration into the Bayesian estimate of the AUV state, the uncertainty of the transform must be modelled, since it represents the weight of the measurement.…”
Section: Introductionmentioning
confidence: 99%
“…From left to right: a submap is fed into PointNetKL to produce a global feature vector that is then fed into a multilayer perceptron to produce the parameters of a Cholesky decomposition and construct a positive-definite covariance matrix. and applying registration methods to these measurements is a well-studied problem [6]. However, when fusing the output of the registration into the Bayesian estimate of the AUV state, the uncertainty of the transform must be modelled, since it represents the weight of the measurement.…”
Section: Introductionmentioning
confidence: 99%