Methods for the Segmentation of Reticular Structures Using 3D LiDAR Data: A Comparative Evaluation
F.J. Soler, A. Peidró, M. Fabregat, L. Paya, O. Reinoso
Computer Modeling in Engineering & Sciences  (June 2025)
Ed. Tech Science Press  ISSN:1526-1506  DOI:http://dx.doi.org/10.32604/cmes.2025.064510  - pp. 1-29

Resumen:

Reticular structures are the basis of major infrastructure projects, including bridges, electrical pylons and airports. However, inspecting and maintaining these structures is both expensive and hazardous, traditionally requiring human involvement.While some research has been conducted in this �eld of study, most ešorts focus on faults identi�cation through images or the design of robotic platforms, oŸen neglecting the autonomous navigation of robots through the structure. is study addresses this limitation by proposing methods to detect navigable surfaces in truss structures, thereby enhancing the autonomous capabilities of climbing robots to navigate through these environments. e paper proposes multiple approaches for the binary segmentation between navigable surfaces and background from 3D point clouds captured from metallic trusses. Approaches can be classi�ed into two paradigms: analytical algorithms and deep learning methods. Within the analytical approach, an ad hoc algorithm is developed for segmenting the structures, leveraging dišerent techniques to evaluate the eigendecomposition of planar patches within the point cloud. In parallel, widely used and advanced deep learning models, including PointNet, PointNet++, MinkUNet34C, and PointTransformerV3, are trained and evaluated for the same task. A comparative analysis of these paradigms reveals some key insights. e analytical algorithm demonstrates easier parameter adjustment and comparable performance to that of the deep learning models, despite the latter’s higher computational demands. Nevertheless, the deep learning models stand out in segmentation accuracy, with PointTransformerV3 achieving impressive results, such as a Mean Intersection Over Union (mIoU) of approximately 97%. is study highlights the potential of analytical and deep learning approaches to improve the autonomous navigation of climbing robots in complex truss structures.e �ndings underscore the trade-ošs between computational e›ciency and segmentation performance, ošering valuable insights for future research and practical applications in autonomous infrastructure maintenance and inspection.