Organized Segmentation Tools is a utility for segmenting organized point clouds, as produced by an RGB-D sensor. This includes plane segmentation, euclidean clustering, and edge detection. The provided library enables these to be run in a multi-threaded pipeline using Intel's Threading Building Blocks (TBB). For example, normal estimation, plane segmentation, and clustering may be run concurrently. The project builds both a shared library suitable for use in other projects, and an example application demonstrating the tools as applied to a live stream from an RGB-D sensor.
mkdir buildcd buildcmake ..make
./organized_segmentation_tbb_demo
The following papers describe the techniques used here:
Efficient Organized Point Cloud Segmentation with Connected Components
@inproceedings{trevor2013efficient, title={Efficient Organized Point Cloud Segmentation with Connected Components}, author={Trevor, A and Gedikli, Suat and Rusu, R and Christensen, H}, booktitle={3rd Workshop on Semantic Perception Mapping and Exploration (SPME), Karlsruhe, Germany}, year={2013} }
RGB-D edge detection and edge-based registration
@inproceedings{choi2013rgb, title={RGB-D edge detection and edge-based registration}, author={Choi, Changhyun and Trevor, Alexander JB and Christensen, Henrik I}, booktitle={Intelligent Robots and Systems (IROS), 2013 IEEE/RSJ International Conference on}, pages={1568--1575}, year={2013}, organization={IEEE} }
@inproceedings{holzer2012adaptive, title={Adaptive neighborhood selection for real-time surface normal estimation from organized point cloud data using integral images}, author={Holzer, Stefan and Rusu, Radu Bogdan and Dixon, M and Gedikli, Suat and Navab, Nassir}, booktitle={Intelligent Robots and Systems (IROS), 2012 IEEE/RSJ International Conference on}, pages={2684--2689}, year={2012}, organization={IEEE} }
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent