Registration of three‐dimensional scanning LiDAR sensors: An evaluation of model‐based and model‐free methods

D’Adamo, Timothy; Phillips, Tyson Govan; McĂree, P. R.

doi:10.1002/rob.21811

Cited by 12 publications

(7 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The uncertainty in the sensor’s pose relates to how accurate the extrinsic calibration or registration procedure is. In previous studies we have estimated that this can be determined within several millimetres/milliradians for translational and rotational components, respectively, [ 13 , 14 ]. Figure 6 illustrates the effect of registration uncertainty,

, on the range measurement distribution,

.…”

Section: The Maximum Sum Of Evidence (Msoe) Methodsmentioning

confidence: 99%

Maximum Sum of Evidence—An Evidence-Based Solution to Object Pose Estimation in Point Cloud Data

Phillips

D’Adamo

McĂree

2021

Sensors

Self Cite

View full text Add to dashboard Cite

The capability to estimate the pose of known geometry from point cloud data is a frequently arising requirement in robotics and automation applications. This problem is directly addressed by Iterative Closest Point (ICP), however, this method has several limitations and lacks robustness. This paper makes the case for an alternative method that seeks to find the most likely solution based on available evidence. Specifically, an evidence-based metric is described that seeks to find the pose of the object that would maximise the conditional likelihood of reproducing the observed range measurements. A seedless search heuristic is also provided to find the most likely pose estimate in light of these measurements. The method is demonstrated to provide for pose estimation (2D and 3D shape poses as well as joint-space searches), object identification/classification, and platform localisation. Furthermore, the method is shown to be robust in cluttered or non-segmented point cloud data as well as being robust to measurement uncertainty and extrinsic sensor calibration.

show abstract

, on the range measurement distribution,

.…”

Section: The Maximum Sum Of Evidence (Msoe) Methodsmentioning

confidence: 99%

Maximum Sum of Evidence—An Evidence-Based Solution to Object Pose Estimation in Point Cloud Data

Phillips

D’Adamo

McĂree

2021

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…If the sensor is installed on a robotic platform, it is usually necessary to know its location relative to a navigation reference frame on that platform and small orientation errors translate to large point cloud errors at long range. Extrinsic registration procedures to determine this have inherent uncertainty [ 7 , 8 ]. The challenge of reliability is magnified in field robotics environments, which are cluttered, unordered, and can be in a constant state of unpredictable flux [ 9 ].…”

Section: The Challenges Of Estimating Object Pose In Point Cloud Datamentioning

confidence: 99%

Real-Time 6-DOF Pose Estimation of Known Geometries in Point Cloud Data

Bhandari

Phillips

McĂree

2023

Sensors

Self Cite

View full text Add to dashboard Cite

The task of tracking the pose of an object with a known geometry from point cloud measurements arises in robot perception. It calls for a solution that is both accurate and robust, and can be computed at a rate that aligns with the needs of a control system that might make decisions based on it. The Iterative Closest Point (ICP) algorithm is widely used for this purpose, but it is susceptible to failure in practical scenarios. We present a robust and efficient solution for pose-from-point cloud estimation called the Pose Lookup Method (PLuM). PLuM is a probabilistic reward-based objective function that is resilient to measurement uncertainty and clutter. Efficiency is achieved through the use of lookup tables, which substitute complex geometric operations such as raycasting used in earlier solutions. Our results show millimetre accuracy and fast pose estimation in benchmark tests using triangulated geometry models, outperforming state-of-the-art ICP-based methods. These results are extended to field robotics applications, resulting in real-time haul truck pose estimation. By utilising point clouds from a LiDAR fixed to a rope shovel, the PLuM algorithm tracks a haul truck effectively throughout the excavation load cycle at a rate of 20 Hz, matching the sensor frame rate. PLuM is straightforward to implement and provides dependable and timely solutions in demanding environments.

show abstract

“…It also covers inevitable errors in the extrinsic calibration parameters, that we used to register the LIDAR and the localization. Indeed, such parameters are only accurate up to a certain point (D'Adamo et al, 2018). Then, if

X_{i}^{l}

belonged to the ground,

P_{R} (X_{i}^{l})

was estimated as follows: If

X_{i}^{m}

falls into a mapped road:

P_{R} (X_{i}^{l}) = \int_{- \infty}^{d_{i}} \frac{1}{σ_{b} \sqrt{2 π}} {exp}^{- \frac{1}{2} {(\frac{x}{σ_{b}})}^{2}} d x

Otherwise:

P_{R} (X_{i}^{l}) = 1 - \int_{- \infty}^{d_{i}} \frac{1}{σ_{b} \sqrt{2 π}} {exp}^{- \frac{1}{2} {(\frac{x}{σ_{b}})}^{2}} d x

It has to be noted that, even though the motion was compensated in the forward direction via Equation (14) this probability is associated to the noncompensated LIDAR point.…”

Section: Automatic Labellisation Procedures Of Lidar Scans From Lane‐level Hd Mapsmentioning

confidence: 99%

Fusion of neural networks, for LIDAR‐based evidential road mapping

Capellier

Davoine

Cherfaoui

et al. 2021

Journal of Field Robotics

View full text Add to dashboard Cite

LIDAR sensors are usually used to provide autonomous vehicles with three‐dimensional representations of their environment. In ideal conditions, geometrical models could detect the road in LIDAR scans, at the cost of a manual tuning of numerical constraints, and a lack of flexibility. We instead propose an evidential pipeline, to accumulate road detection results obtained from neural networks. First, we introduce RoadSeg, a new convolutional architecture that is optimized for road detection in LIDAR scans. RoadSeg is used to classify individual LIDAR points as either belonging to the road, or not. Yet, such point‐level classification results need to be converted into a dense representation, that can be used by an autonomous vehicle. We thus second present an evidential road mapping algorithm, that fuses consecutive road detection results. We benefitted from a reinterpretation of logistic classifiers, which can be seen as generating a collection of simple evidential mass functions. An evidential grid map that depicts the road can then be obtained, by projecting the classification results from RoadSeg into grid cells, and by handling moving objects via conflict analysis. The system was trained and evaluated on real‐life data. A python implementation maintains a 10 Hz framerate. Since road labels were needed for training, a soft labeling procedure, relying lane‐level HD maps, was used to generate coarse training and validation sets. An additional test set was manually labeled for evaluation purposes. So as to reach satisfactory results, the system fuses road detection results obtained from three variants of RoadSeg, processing different LIDAR features.

show abstract

Registration of three‐dimensional scanning LiDAR sensors: An evaluation of model‐based and model‐free methods

Cited by 12 publications

References 25 publications

Maximum Sum of Evidence—An Evidence-Based Solution to Object Pose Estimation in Point Cloud Data

Maximum Sum of Evidence—An Evidence-Based Solution to Object Pose Estimation in Point Cloud Data

Real-Time 6-DOF Pose Estimation of Known Geometries in Point Cloud Data

Fusion of neural networks, for LIDAR‐based evidential road mapping

Contact Info

Product

Resources

About