Fast Dual Decomposition based Mesh-Graph Clustering for Point Clouds

Burger, Patrick; Naujoks, Benjamin; Wuensche, Hans-Joachim

doi:10.1109/itsc.2018.8569999

Cited by 13 publications

(8 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the face of different application requirements, we need to choose some suitable segmentation standards. The standard method is to use Euclidean distance and segments according to the distribution characteristics of continuous points or to calculate the angle relationship of adjacent point clouds [13,15], and an optimal neighborhood should be selected [40]. Considering the difference in point cloud density, we design different segmentation standards in horizontal and vertical directions.…”

Section: Analysis Of Clustering Tasksmentioning

confidence: 99%

“…The method to update the category greatly affects the real-time performance of the point cloud clustering. Patrick Burger [15] updates the categories after the initial segmentation using connection maps in different directions. It is a common practice to use a breadth-first-search algorithm in the search, which is used in many fields of automatic vehicles [13,41].…”

Section: Analysis Of Clustering Tasksmentioning

confidence: 99%

“…Patrick Burger proposed a mesh-graph based clustering. It traverses the point cloud from different directions, integrates multiple features as the evaluation criteria, and uses the structured neighborhood relationship between point clouds to improve the operation efficiency [15,16]. These methods have made a lot of contributions to the point cloud storage structure.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Two-Layer-Graph Clustering for Real-Time 3D LiDAR Point Cloud Segmentation

et al. 2020

View full text Add to dashboard Cite

The perception system has become a topic of great importance for autonomous vehicles, as high accuracy and real-time performance can ensure safety in complex urban scenarios. Clustering is a fundamental step for parsing point cloud due to the extensive input data (over 100,000 points) of a wide variety of complex objects. It is still challenging to achieve high precision real-time performance with limited vehicle-mounted computing resources, which need to balance the accuracy and processing time. We propose a method based on a Two-Layer-Graph (TLG) structure, which can be applied in a real autonomous vehicle under urban scenarios. TLG can describe the point clouds hierarchically, we use a range graph to represent point clouds and a set graph for point cloud sets, which reduce both processing time and memory consumption. In the range graph, Euclidean distance and the angle of the sensor position with two adjacent vectors (calculated from continuing points to different direction) are used as the segmentation standard, which use the local concave features to distinguish different objects close to each other. In the set graph, we use the start and end position to express the whole set of continuous points concisely, and an improved Breadth-First-Search (BFS) algorithm is designed to update categories of point cloud sets between different channels. This method is evaluated on real vehicles and major datasets. The results show that TLG succeeds in providing a real-time performance (less than 20 ms per frame), and a high segmentation accuracy rate (93.64%) for traffic objects in the road of urban scenarios.

show abstract

Section: Analysis Of Clustering Tasksmentioning

confidence: 99%

Section: Analysis Of Clustering Tasksmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Two-Layer-Graph Clustering for Real-Time 3D LiDAR Point Cloud Segmentation

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Furthermore, we compare against a simple point tracking approach (PT) [26], where the bounding box center as well as their dimensions are estimated. The segmented point clouds of the target vehicle are provided from the method of [27], [28]. During the scenario, different occlusions (only the back or one side can be seen) of the segmented vehicle and segmentation errors, e. g., falsely associated ground plane points, occur.…”

Section: B Nurbs Shape Functionmentioning

confidence: 99%

Fast 3D Extended Target Tracking using NURBS Surfaces

Naujoks

Burger

Wuensche

2019

2019 IEEE Intelligent Transportation Systems Conference (ITSC)

Self Cite

View full text Add to dashboard Cite

This paper proposes fast and novel methods to jointly estimate the target's unknown 3D shape and dynamics. Measurements are noisy and sparsely distributed 3D points from a light detection and ranging (LiDAR) sensor. The methods utilize non-uniform rational B-splines (NURBS) surfaces to approximate the target's shape. One method estimates Cartesian scaling parameters of a NURBS surface, whereas the second method estimates the corresponding NURBS weights, too. Major advantages are the capability of estimating a fully 3D shape as well as the fast processing time. Real-world evaluations with a static and dynamic vehicle show promising results compared to state-of-the-art 3D extended target tracking algorithms.

show abstract

“…compare our results with the well-known Labeled Multi-Bernoulli Filter (LMB) [4], its variation the Generalized-LMB (GLMB) [20] and a classical filtering approach with an underlying track management [21] denoted as BuTd. As the other filters need clustered detections, we cluster the point cloud in coherent objects with the methods of [9], [10], [22]. Then, the x and y positions of the bounding box centers z x i and z y i are used as measurements for the different methods.…”

Section: (B)mentioning

confidence: 99%

The Greedy Dirichlet Process Filter -An Online Clustering Multi-Target Tracker

Naujoks

Burger

Wuensche

2018

2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)

Self Cite

View full text Add to dashboard Cite

Reliable collision avoidance is one of the main requirements for autonomous driving. Hence, it is important to correctly estimate the states of an unknown number of static and dynamic objects in real-time. Here, data association is a major challenge for every multi-target tracker. We propose a novel multi-target tracker called Greedy Dirichlet Process Filter (GDPF) based on the non-parametric Bayesian model called Dirichlet Processes and the fast posterior computation algorithm Sequential Updating and Greedy Search (SUGS). By adding a temporal dependence we get a real-time capable tracking framework without the need of a previous clustering or data association step. Real-world tests show that GDPF outperforms other multi-target tracker in terms of accuracy and stability.

show abstract

Fast Dual Decomposition based Mesh-Graph Clustering for Point Clouds

Cited by 13 publications

References 13 publications

Two-Layer-Graph Clustering for Real-Time 3D LiDAR Point Cloud Segmentation

Two-Layer-Graph Clustering for Real-Time 3D LiDAR Point Cloud Segmentation

Fast 3D Extended Target Tracking using NURBS Surfaces

The Greedy Dirichlet Process Filter -An Online Clustering Multi-Target Tracker

Contact Info

Product

Resources

About