A Review of Point Cloud Registration Algorithms for Mobile Robotics

By François Pomerleau, Francis Colas, Roland Yves Siegwart. Found. Trends Robotics, 2015

772 citations recorded by the source corpus.

The topic of this review is geometric registration in robotics. Registrationalgorithms associate sets of data into a common coordinate system.They have been used extensively in object reconstruction, inspection,medical application, and localization of mobile robotics. We focus onmobile robotics applications in which point clouds are to be registered.While the underlying principle of those algorithms is simple, manyvariations have been proposed for many different applications. In thisreview, we give a historical perspective of the registration problem andshow that the plethora of solutions can be organized and differentiatedaccording to a few elements. Accordingly, we present a formalizationof geometric registration and cast algorithms proposed in the literatureinto this framework. Finally, we review a few applications of thisframework in mobile robotics that cover different kinds of platforms,environments, and tasks. These examples allow us to study the specificrequirements of each use case and the necessary configuration choicesleading to the registration implementation. Ultimately, the objective ofthis review is to provide guidelines for the choice of geometric registrationconfiguration.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Mechanical Engineering

View Paper: https://www.semanticscholar.org/paper/16af1ee91c81caaba4c941c39a5aa99818188355

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Mechanical Engineering

Citations in the corpus

The current state and future outlook of rescue robotics (2019, background): “Many of these advances, for example point cloud registration for mobile robot localization and mapping (Pomerleau et al., 2015; Dubé et al., 2016) and autonomous stair climbing (Ohashi et al.”
Review: deep learning on 3D point clouds (2020, background): “Point clouds have a range of applications in different areas such as robotics [14], autonomous driving [15], augmented and virtual reality [16], manufacturing and building rendering [17], etc.”
LINS: A Lidar-Inertial State Estimator for Robust and Efficient Navigation (2019, background): “In iterated Kalman filtering, the state update can be linked to an optimization problem [7], [29] considering the deviation from the prior −xk bk+1 and the residual functions, f (·), (i.”
HDMapNet: An Online HD Map Construction and Evaluation Framework (2021, methodology): “First, pairwise alignment algorithm like ICP [1], NDT [2] and their variants [3] are employed to match LiDAR data at two nearby timestamps using semantic [4] or geometry information [5].”
Deep Closest Point: Learning Representations for Point Cloud Registration (2019, methodology): “For more information, [32, 37] summarize ICP and its variants developed over the last 20 years.”
Feature-Metric Registration: A Fast Semi-Supervised Approach for Robust Point Cloud Registration Without Correspondences (2020, methodology): “Before deep learning is prevalent, most of point cloud registration methods [24, 15, 19, 3] use two-stage optimisation-based algorithms: correspondence searching and transformation estimation.”
IMLS-SLAM: Scan-to-Model Matching Based on 3D Data (2018, background): “[1] and more recently [2] have surveyed efficient variants of ICP, such as the point-to-plane matching.”
PointDSC: Robust Point Cloud Registration using Deep Spatial Consistency (2021, background): “, [8, 1, 47, 31, 44, 44]) have been comprehensively reviewed in [52].”
PRNet: Self-Supervised Learning for Partial-to-Partial Registration (2019, methodology): “ICP [5] and variants [6, 7, 8, 9] have been widely used for registration.”
A comprehensive survey on point cloud registration (2021, methodology): “Recently, [78] proposes a review of point cloud registration in robotics.”