Small Object Tracking in LiDAR Point Clouds: Learning the Target-Awareness Prototype and Fine-Grained Search Region
Light Detection and Ranging (LiDAR) point clouds are an essential perception modality for artificial intelligence systems like autonomous driving and robotics, where the ubiquity of small objects in real-world scenarios substantially challenges the visual tracking of small targets amidst the vastnes...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/12/3633 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839652713879568384 |
|---|---|
| author | Shengjing Tian Yinan Han Xiantong Zhao Xiuping Liu |
| author_facet | Shengjing Tian Yinan Han Xiantong Zhao Xiuping Liu |
| author_sort | Shengjing Tian |
| collection | DOAJ |
| description | Light Detection and Ranging (LiDAR) point clouds are an essential perception modality for artificial intelligence systems like autonomous driving and robotics, where the ubiquity of small objects in real-world scenarios substantially challenges the visual tracking of small targets amidst the vastness of point cloud data. Current methods predominantly focus on developing universal frameworks for general object categories, often sidelining the persistent difficulties associated with small objects. These challenges stem from a scarcity of foreground points and a low tolerance for disturbances. To this end, we propose a deep neural network framework that trains a Siamese network for feature extraction and innovatively incorporates two pivotal modules: the target-awareness prototype mining (TAPM) module and the regional grid subdivision (RGS) module. The TAPM module utilizes the reconstruction mechanism of the masked auto-encoder to distill prototypes within the feature space, thereby enhancing the salience of foreground points and aiding in the precise localization of small objects. To heighten the tolerance of disturbances in feature maps, the RGS module is devised to retrieve detailed features of the search area, capitalizing on Vision Transformer and pixel shuffle technologies. Furthermore, beyond standard experimental configurations, we have meticulously crafted scaling experiments to assess the robustness of various trackers when dealing with small objects. Comprehensive evaluations show our method achieves a mean Success of 64.9% and 60.4% under original and scaled settings, outperforming benchmarks by +3.6% and +5.4%, respectively. |
| format | Article |
| id | doaj-art-571937a2c83d4f2d98e2f35a8b5b01e5 |
| institution | Matheson Library |
| issn | 1424-8220 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-571937a2c83d4f2d98e2f35a8b5b01e52025-06-25T14:25:19ZengMDPI AGSensors1424-82202025-06-012512363310.3390/s25123633Small Object Tracking in LiDAR Point Clouds: Learning the Target-Awareness Prototype and Fine-Grained Search RegionShengjing Tian0Yinan Han1Xiantong Zhao2Xiuping Liu3School of Economics and Management, China University of Mining and Technology, Xuzhou 221116, ChinaDUT-BSU Joint Institute, Dalian University of Technology, Dalian 116024, ChinaSchool of Mathematical Sciences, Dalian University of Technology, Dalian 116024, ChinaDUT-BSU Joint Institute, Dalian University of Technology, Dalian 116024, ChinaLight Detection and Ranging (LiDAR) point clouds are an essential perception modality for artificial intelligence systems like autonomous driving and robotics, where the ubiquity of small objects in real-world scenarios substantially challenges the visual tracking of small targets amidst the vastness of point cloud data. Current methods predominantly focus on developing universal frameworks for general object categories, often sidelining the persistent difficulties associated with small objects. These challenges stem from a scarcity of foreground points and a low tolerance for disturbances. To this end, we propose a deep neural network framework that trains a Siamese network for feature extraction and innovatively incorporates two pivotal modules: the target-awareness prototype mining (TAPM) module and the regional grid subdivision (RGS) module. The TAPM module utilizes the reconstruction mechanism of the masked auto-encoder to distill prototypes within the feature space, thereby enhancing the salience of foreground points and aiding in the precise localization of small objects. To heighten the tolerance of disturbances in feature maps, the RGS module is devised to retrieve detailed features of the search area, capitalizing on Vision Transformer and pixel shuffle technologies. Furthermore, beyond standard experimental configurations, we have meticulously crafted scaling experiments to assess the robustness of various trackers when dealing with small objects. Comprehensive evaluations show our method achieves a mean Success of 64.9% and 60.4% under original and scaled settings, outperforming benchmarks by +3.6% and +5.4%, respectively.https://www.mdpi.com/1424-8220/25/12/3633point cloudsLiDARsmall objectsvisual trackingdeep neural network |
| spellingShingle | Shengjing Tian Yinan Han Xiantong Zhao Xiuping Liu Small Object Tracking in LiDAR Point Clouds: Learning the Target-Awareness Prototype and Fine-Grained Search Region Sensors point clouds LiDAR small objects visual tracking deep neural network |
| title | Small Object Tracking in LiDAR Point Clouds: Learning the Target-Awareness Prototype and Fine-Grained Search Region |
| title_full | Small Object Tracking in LiDAR Point Clouds: Learning the Target-Awareness Prototype and Fine-Grained Search Region |
| title_fullStr | Small Object Tracking in LiDAR Point Clouds: Learning the Target-Awareness Prototype and Fine-Grained Search Region |
| title_full_unstemmed | Small Object Tracking in LiDAR Point Clouds: Learning the Target-Awareness Prototype and Fine-Grained Search Region |
| title_short | Small Object Tracking in LiDAR Point Clouds: Learning the Target-Awareness Prototype and Fine-Grained Search Region |
| title_sort | small object tracking in lidar point clouds learning the target awareness prototype and fine grained search region |
| topic | point clouds LiDAR small objects visual tracking deep neural network |
| url | https://www.mdpi.com/1424-8220/25/12/3633 |
| work_keys_str_mv | AT shengjingtian smallobjecttrackinginlidarpointcloudslearningthetargetawarenessprototypeandfinegrainedsearchregion AT yinanhan smallobjecttrackinginlidarpointcloudslearningthetargetawarenessprototypeandfinegrainedsearchregion AT xiantongzhao smallobjecttrackinginlidarpointcloudslearningthetargetawarenessprototypeandfinegrainedsearchregion AT xiupingliu smallobjecttrackinginlidarpointcloudslearningthetargetawarenessprototypeandfinegrainedsearchregion |