Open3D: A Modern Library for 3D Data Processing
About
Open3D is an open-source library that supports rapid development of software that deals with 3D data. The Open3D frontend exposes a set of carefully selected data structures and algorithms in both C++ and Python. The backend is highly optimized and is set up for parallelization. Open3D was developed from a clean slate with a small and carefully considered set of dependencies. It can be set up on different platforms and compiled from source with minimal effort. The code is clean, consistently styled, and maintained via a clear code review mechanism. Open3D has been used in a number of published research projects and is actively deployed in the cloud. We welcome contributions from the open-source community.
Qian-Yi Zhou, Jaesik Park, Vladlen Koltun• 2018
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Point cloud registration | ModelNet40 (Unseen categories) | RMSE (Rotation)29.8764 | 36 | |
| Point cloud registration | ModelNet40 (test) | Inference Time (s)0.004 | 24 | |
| 3D Point Cloud Registration | 3DMatch | Translation Error (cm)8.85 | 20 | |
| Geometric Registration | KITTI Dataset (test) | -- | 14 | |
| 6D Object Pose Tracking | NOCS (test) | 5°5cm Accuracy16.9 | 8 | |
| Multi-way Registration | Augmented ICL-NUIM Living room 2 simulated depth noise (test) | ATE (cm)19.33 | 6 | |
| Multi-way Registration | Augmented ICL-NUIM Office 1 simulated depth noise (test) | ATE (cm)14.42 | 6 | |
| Multi-way Registration | Augmented ICL-NUIM Office 2 simulated depth noise (test) | ATE (cm)17.31 | 6 | |
| Point cloud registration | ModelNet40 v1 (test) | MSE (R)894.9 | 6 | |
| Point cloud registration | ModelNet40 with Gaussian noise (test) | MSE (R)882.6 | 6 |
Showing 10 of 29 rows