Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data
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 de...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-05-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/10/3410 |
| _version_ | 1797534162073157632 |
|---|---|
| author | Claudia Malzer Marcus Baum |
| author_facet | Claudia Malzer Marcus Baum |
| author_sort | Claudia Malzer |
| collection | DOAJ |
| description | 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. |
| first_indexed | 2024-03-10T11:25:41Z |
| format | Article |
| id | doaj.art-4381c84b91c04dbda8d15422381b74a3 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T11:25:41Z |
| publishDate | 2021-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-4381c84b91c04dbda8d15422381b74a32023-11-21T19:39:06ZengMDPI AGSensors1424-82202021-05-012110341010.3390/s21103410Constraint-Based Hierarchical Cluster Selection in Automotive Radar DataClaudia Malzer0Marcus Baum1Data Fusion Group, Institute of Computer Science, University of Göttingen, 37077 Göttingen, GermanyData Fusion Group, Institute of Computer Science, University of Göttingen, 37077 Göttingen, GermanyHigh-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.https://www.mdpi.com/1424-8220/21/10/3410hierarchical clusteringHDBSCANconstraint-based clusteringsemi-supervised clusteringautomotive radar |
| spellingShingle | Claudia Malzer Marcus Baum Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data Sensors hierarchical clustering HDBSCAN constraint-based clustering semi-supervised clustering automotive radar |
| title | Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data |
| title_full | Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data |
| title_fullStr | Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data |
| title_full_unstemmed | Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data |
| title_short | Constraint-Based Hierarchical Cluster Selection in Automotive Radar Data |
| title_sort | constraint based hierarchical cluster selection in automotive radar data |
| topic | hierarchical clustering HDBSCAN constraint-based clustering semi-supervised clustering automotive radar |
| url | https://www.mdpi.com/1424-8220/21/10/3410 |
| work_keys_str_mv | AT claudiamalzer constraintbasedhierarchicalclusterselectioninautomotiveradardata AT marcusbaum constraintbasedhierarchicalclusterselectioninautomotiveradardata |