An Adaptive 3D Grid-Based Clustering Algorithm for Automotive High Resolution Radar Sensor

Li, Mingkang; Stolz, Martin; Feng, Zhaofei; Kunert, Martin; Henze, Roman; Küçükay, Ferit

doi:10.1109/icves.2018.8519483

Cited by 25 publications

(14 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Grid-based DBSCAN algorithm uses

grid modeling to solve the clustering difficulties caused by low angular resolution [ 7 , 30 ]. On this basis, Doppler velocity is added to help improve the clustering effect and adaptive clustering method for tracking is introduced to further enhance algorithm realizability [ 8 ]. The methods listed above are suitable for high-resolution radar data.…”

Section: Data Models and Representations From Mmw Radarmentioning

confidence: 99%

“…Therefore, original point cloud information before clustering and tracking which is called cluster-layer data is used more frequently at high-level automated driving. In these applications, raw point cloud data of single snapshot is used to obtan object dimension [ 7 , 8 ], orientation, motion estimation [ 9 , 10 ], and object category [ 11 , 12 ]. Then, raw radar data accumulated from multiple snapshots is used to build grid maps [ 13 , 14 ].…”

Section: Introductionmentioning

confidence: 99%

“…Then, raw radar data accumulated from multiple snapshots is used to build grid maps [ 13 , 14 ]. These representations are used to express dynamic and static environment elements and applied to many applications such as object detection and tracking [ 8 , 15 , 16 ], environment mapping, and vehicle localization [ 17 , 18 , 19 ].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

MMW Radar-Based Technologies in Autonomous Driving: A Review

Zhou

Yang

Jiang

et al. 2020

Sensors

View full text Add to dashboard Cite

With the rapid development of automated vehicles (AVs), more and more demands are proposed towards environmental perception. Among the commonly used sensors, MMW radar plays an important role due to its low cost, adaptability In different weather, and motion detection capability. Radar can provide different data types to satisfy requirements for various levels of autonomous driving. The objective of this study is to present an overview of the state-of-the-art radar-based technologies applied In AVs. Although several published research papers focus on MMW Radars for intelligent vehicles, no general survey on deep learning applied In radar data for autonomous vehicles exists. Therefore, we try to provide related survey In this paper. First, we introduce models and representations from millimeter-wave (MMW) radar data. Secondly, we present radar-based applications used on AVs. For low-level automated driving, radar data have been widely used In advanced driving-assistance systems (ADAS). For high-level automated driving, radar data is used In object detection, object tracking, motion prediction, and self-localization. Finally, we discuss the remaining challenges and future development direction of related studies.

show abstract

“…Grid-based DBSCAN algorithm uses

Section: Data Models and Representations From Mmw Radarmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

MMW Radar-Based Technologies in Autonomous Driving: A Review

Zhou

Yang

Jiang

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…While HDBSCAN [ 4 ], a hierarchical DBSCAN variant and current state of the art in hierarchical clustering, has gained a lot of attention across many different research fields in recent years, it has rarely been used in the context of radar data. While some authors did not consider hierarchical methods because of their high runtime complexity compared to non-hierarchical methods [ 1 , 5 ], Schumann et al [ 2 ] tested HDBSCAN on radar measurements of stationary and moving objects but found it not suitable for their task.…”

Section: Introductionmentioning

confidence: 99%

Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data

Malzer

Baum

2021

Sensors

View full text Add to dashboard Cite

High-resolution automotive radar sensors play an increasing role in detection, classification and tracking of moving objects in traffic scenes. Clustering is frequently used to group detection points in this context. However, this is a particularly challenging task due to variations in number and density of available data points across different scans. Modified versions of the density-based clustering method DBSCAN have mostly been used so far, while hierarchical approaches are rarely considered. In this article, we explore the applicability of HDBSCAN, a hierarchical DBSCAN variant, for clustering radar measurements. To improve results achieved by its unsupervised version, we propose the use of cluster-level constraints based on aggregated background information from cluster candidates. Further, we propose the application of a distance threshold to avoid selection of small clusters at low hierarchy levels. Based on exemplary traffic scenes from nuScenes, a publicly available autonomous driving data set, we test our constraint-based approach along with other methods, including label-based semi-supervised HDBSCAN. Our experiments demonstrate that cluster-level constraints help to adjust HDBSCAN to the given application context and can therefore achieve considerably better results than the unsupervised method. However, the approach requires carefully selected constraint criteria that can be difficult to choose in constantly changing environments.

show abstract

“…The clustering of radar data is usually done either in the Cartesian domain, e.g., with a dynamically self-configuring approach based on the density of the cluster [8] or in the range-DoA domain to accommodate for the operating principle of radar data [9]. This modification is extended in [10] to include the dimension of velocity and [11] proposes a direct application on the measured data in range, velocity, and direction of arrival (DoA), with different radii for each domain dependent on the radar system parameters.…”

Section: Introductionmentioning

confidence: 99%

Clustering of Closely Adjacent Extended Objects in Radar Images using Velocity Profile Analysis

Schlichenmaier¹,

Roos²,

Hügler³

et al. 2019

2019 IEEE MTT-S International Conference on Microwaves for Intelligent Mobility (ICMIM)

View full text Add to dashboard Cite

As high resolution automotive radars become more common, so does their usage for next-generation functionalities like advanced driver assistant systems and autonomous driving. This creates the need for robust clustering techniques to distinguish among multiple extended objects like vehicles in the same scenario. One especially challenging scenario is that of separating two extended targets close to each other, each following its own trajectory. This paper proposes a clustering algorithm based on the analysis of the velocity profile to divide target points of multiple vehicles into sub-clusters. The theoretical background is explained and shown on simulation data. The algorithm is verified using radar measurements of two extended vehicular targets.

show abstract

An Adaptive 3D Grid-Based Clustering Algorithm for Automotive High Resolution Radar Sensor

Cited by 25 publications

References 17 publications

MMW Radar-Based Technologies in Autonomous Driving: A Review

MMW Radar-Based Technologies in Autonomous Driving: A Review

Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data

Clustering of Closely Adjacent Extended Objects in Radar Images using Velocity Profile Analysis

Contact Info

Product

Resources

About