A Storage-Efficient Feature for 3D Concrete Defect Segmentation to Replace Normal Vector
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909922437890048 |
|---|---|
| author | Hua, Linxin Deng, Jianghua Lu, Ye |
| author_facet | Hua, Linxin Deng, Jianghua Lu, Ye |
| contents | Point cloud reconstruction of damage offers an effective solution to image-based methods vulnerable to background noise, yet its application is constrained by the high volume of 3D data. This study proposes a new feature, relative angle, computed as the angle between the normal vector of a point and the average normal vector of its parent point cloud. This single-dimensional feature provides directionality information equivalent to normal vectors for concrete surface defect characteristics. Through entropy-based feature evaluation, this study demonstrates the ability of relative angle to filter out redundant information in undamaged sections while retaining effective information in damaged sections. By training and testing with PointNet++, models based on the relative angles achieved similar performance to that of models based on normal vectors while delivering 27.6% storage reduction and 83% input channel compression. This novel feature has the potential to enable larger-batch execution on resource-constrained hardware without the necessity of architectural modifications to models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_19760 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Storage-Efficient Feature for 3D Concrete Defect Segmentation to Replace Normal Vector Hua, Linxin Deng, Jianghua Lu, Ye Computer Vision and Pattern Recognition I.4.6; I.5.3; J.6 Point cloud reconstruction of damage offers an effective solution to image-based methods vulnerable to background noise, yet its application is constrained by the high volume of 3D data. This study proposes a new feature, relative angle, computed as the angle between the normal vector of a point and the average normal vector of its parent point cloud. This single-dimensional feature provides directionality information equivalent to normal vectors for concrete surface defect characteristics. Through entropy-based feature evaluation, this study demonstrates the ability of relative angle to filter out redundant information in undamaged sections while retaining effective information in damaged sections. By training and testing with PointNet++, models based on the relative angles achieved similar performance to that of models based on normal vectors while delivering 27.6% storage reduction and 83% input channel compression. This novel feature has the potential to enable larger-batch execution on resource-constrained hardware without the necessity of architectural modifications to models. |
| title | A Storage-Efficient Feature for 3D Concrete Defect Segmentation to Replace Normal Vector |
| topic | Computer Vision and Pattern Recognition I.4.6; I.5.3; J.6 |
| url | https://arxiv.org/abs/2511.19760 |