3D Understanding of Deformable Linear Objects: Datasets and Transferability Benchmark

Bare Luka Žagar,Tim Hertel,Mingyu Liu,Ekim Yurtsever,ALois C. Knoll

2023-10-13

Abstract:Deformable linear objects are vastly represented in our everyday lives. It is often challenging even for humans to visually understand them, as the same object can be entangled so that it appears completely different. Examples of deformable linear objects include blood vessels and wiring harnesses, vital to the functioning of their corresponding systems, such as the human body and a vehicle. However, no point cloud datasets exist for studying 3D deformable linear objects. Therefore, we are introducing two point cloud datasets, PointWire and PointVessel. We evaluated state-of-the-art methods on the proposed large-scale 3D deformable linear object benchmarks. Finally, we analyzed the generalization capabilities of these methods by conducting transferability experiments on the PointWire and PointVessel datasets.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The paper aims to address the understanding and analysis of Deformable Linear Objects (DLOs) in 3D space. Specifically, it focuses on the topological understanding and untangling of complex deformable linear structures such as blood vessels and wire harnesses. The current issue is the lack of 3D point cloud datasets for these complex DLOs, especially in the fields of automated wire harness processing and medical imaging. Therefore, the authors propose two new point cloud datasets—PointWire and PointVessel.

- The **PointWire** dataset is based on real automotive wire harness scans and is expanded to a large-scale sample through a semi-automated data generation pipeline, aiming to promote the development of wire harness automation research.
- The **PointVessel** dataset comes from synthetic medical vessel data, converted into point cloud format to support the study of vascular topology.

In addition, the paper evaluates the performance of several existing advanced point cloud segmentation methods on these newly proposed benchmark datasets and conducts transferability experiments to verify whether these methods can learn general topological concepts between different types of DLOs (such as wire harnesses and vascular systems). The experimental results show that although some methods perform well on specific tasks, overall, the topological structure segmentation of DLOs remains a challenging task.