How do ICP variants perform when used for scan matching terrain point clouds?

Donoso, Francisco; Austin, Kevin J.; McĂree, P. R.

doi:10.1016/j.robot.2016.10.011

Cited by 38 publications

(37 citation statements)

References 37 publications

(49 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ICP method generates a correspondence between points and finds the rotation and translation that minimizes the distance between all correspondences. This research is extended to generate a correspondence between lines and faces [24], [25]; improved studies with many different versions exist [26]. NDT divides the standard point cloud into equal cell sizes and calculates the normal distribution.…”

Section: B Estimation Of the Relative Pose Scan-matchingmentioning

confidence: 99%

Robust 3-Dimension Point Cloud Mapping in Dynamic Environment Using Point-Wise Static Probability-Based NDT Scan-Matching

et al. 2020

View full text Add to dashboard Cite

Graph-based simultaneous localization and mapping (SLAM) is one of the methods to generate point cloud maps which are used for various applications in autonomous vehicles. Graph-based SLAM represents the pose of the vehicle as a node and the odometry between two different nodes as an edge. Among the edge generating methods, scan matching, light detection and ranging (LiDAR) based method, can provide an accurate pose between two nodes based on the high distance accuracy of the LiDAR. However, the point cloud in real driving situations contains numerous moving objects, which degrade the scan-matching performance. Therefore, this paper defines the static probability which means the likelihood that an acquired point is from a static object, and proposes the weighted normal distribution transformation (NDT), which is achieved by modifying NDT. Weighted NDT is a scan-matching algorithm which can reflect the static probability of each point as a weight. The odometry from the weighted NDT is utilized for graph construction to generate a robust point cloud map even in a dynamic environment. Finally, the proposed algorithm was compared with the existing object removal algorithms in two areas: dynamic object classification and scan-matching performance. Based on the scan-matching results, the accuracy of the point cloud map generated by the proposed algorithm was evaluated with a reference map using highperformance global navigation satellite system (GNSS). It was confirmed that the proposed algorithm has higher classification accuracy and lower scan-matching error compared with other dynamic object removal methods. The proposed algorithm was able to generate a point cloud map, despite the presence of many dynamic objects, that was similar to a map generated in the absence of dynamic objects in the same environment. INDEX TERMS Static probability, Scan-matching, Weighted NDT, LiDAR characteristic, Graph-based SLAM I. INTRODUCTION Light detection and ranging (LiDAR), which is an essential sensor in autonomous vehicles, has the following two characteristics. First, it can provide precise distance information on vehicles, buildings, and people in the form of points. Second, a LiDAR sensor can represent the surroundings of

show abstract

Section: B Estimation Of the Relative Pose Scan-matchingmentioning

confidence: 99%

Robust 3-Dimension Point Cloud Mapping in Dynamic Environment Using Point-Wise Static Probability-Based NDT Scan-Matching

et al. 2020

View full text Add to dashboard Cite

show abstract

“…For a comprehensive breakdown of the ICP process, readers should refer to the work of Rusinkiewicz and Leroy [15]. For further reading on the practical application of ICP, the works of Pomerleau et al [16] and Donoso et al [17] are good starting points.…”

Section: Point Cloud Registration With Iterative Closest Pointmentioning

confidence: 99%

Loss of Significance and Its Effect on Point Normal Orientation and Cloud Registration

et al. 2019

View full text Add to dashboard Cite

Point normal calculation and cloud registration are two of the most common operations in point cloud processing. However, both are vulnerable to issues of numerical precision and loss of significance. This paper documents how loss of significance in the open-source Point Cloud Library can create erroneous point normals and cause cloud registration to fail. Several test clouds are used to demonstrate how the loss of significance is caused by tight point spacing and clouds being shifted far from the origin of their coordinate system. The results show that extreme loss of significance can cause point normals to be calculated with a random orientation, and cause meters of error during cloud registration. Depending on the structure of the point cloud, loss of significance can occur when the cloud is at hundreds or even tens of meters from the origin of its coordinate system. Shifting to larger data types (e.g., from 32-bit “floats” to 64-bit “doubles”) can alleviate the problem but will not solve it completely. Several “best practice” recommendations for avoiding this issue are proposed. But the only solution guaranteed to eliminate loss of significance is de-meaning the entire cloud, or clusters of points before processing.

show abstract

“…Ze-Su et al [35] believe this would represent two O(n) searches: one for the translation estimation, and one for the orientation estimation. This approach is found to be more precise and eicient than ICP in the given examples [35], although given the variation in performance of algorithms in scenes [12] this result may not be generalisable. As demonstrated by Ze-Su et al, the method is applied to identify two complete sets of data, rather than mapping a subset of the data (the area visible around the robot) into the full set of data (the full map); further adaptation may therefore be required for the method to function for general indoor localisation.…”

Section: Existing Workmentioning

confidence: 99%

“…The problem of line-of-sight indoor localisation was irst resolved through the matching of cloud point data (obtained from a line-ofsight sensor such as a Li-Dar) to retrieve tuple (x, y, θ ) describing the location and orientation of a robot in a known environment. This was irst achieved by algorithms such as the Iterative Closest Point (ICP) algorithm Besl and McKay [4], and a long line of alternative heuristic algorithms Lu and Milios [23], Diosi and Kleeman [11] [29], Biber and Strasser [5], Donoso-Aguirre et al [13], Konecny et al [19] and various improvements on the ICP's convergence speed [12] [32], dataset optimisation [33] [25] or precision metrics [14]. Performing indoor localisation without a priori knowledge of the robot's pose increases the diiculty to this problem, as a global search for the position must now be performed, rather than simply a pose reinement.…”

Section: Introductionmentioning

confidence: 99%

Analysis of the use of genetic algorithms for indoor localisation via cloud point matching

Boland

Indrusiak

2018

Proceedings of the Genetic and Evolutionary Computation Conference

View full text Add to dashboard Cite

A system's ability to precisely locate itself in a known physical environment is key to its capacity to interact with the environment in an intricate manner. The indoor localisation problem has been approached in a variety of ways, ranging from the identiication of pre-deined features or topologies to the more general cloud-point matching.Cloud point matching can be achieved using a variety of algorithms, each with beneits and drawbacks. Recent improvements have focused on the application of genetic algorithms to solve the initial 'global' search for a solution, before reining this solution to a precise position through a non-genetic algorithm. This project aims to demonstrate the ineiciency of genetic algorithms applied to the global search problem for the issue of indoor localisation; this is thought to be caused by the solution space's low dimensionality, solution landscape topology and the ineicacy of crossover operators in the algorithm. Based on our assumptions of map topologies, we conclude that signiicant redundancies can be found in some purely genetic heuristics and suggest further development of landscape analysis to allow the use of algorithms appropriate to the scenario's complexity.

show abstract

How do ICP variants perform when used for scan matching terrain point clouds?

Cited by 38 publications

References 37 publications

Robust 3-Dimension Point Cloud Mapping in Dynamic Environment Using Point-Wise Static Probability-Based NDT Scan-Matching

Robust 3-Dimension Point Cloud Mapping in Dynamic Environment Using Point-Wise Static Probability-Based NDT Scan-Matching

Loss of Significance and Its Effect on Point Normal Orientation and Cloud Registration

Analysis of the use of genetic algorithms for indoor localisation via cloud point matching

Contact Info

Product

Resources

About