Point Cloud segmentation techinique based on Watershed Transform implemented with IFT (seeds from Region Growing)
| Build Platform | Status |
|---|---|
| Ubuntu | ci-ubuntu-16.04 ci-ubuntu-18.04 |
- GCC 7.1 (supports OpenMP 4.5) *needed for parallelism
- CMake 3.8
- PCL 1.8
- OpenCV 3.2
Please cite these papers in your publications if it helps your research.
@article{paiva2020historical, title={Historical Building Point Cloud Segmentation Combining Hierarchical Watershed Transform and Curvature Analysis}, author={Paiva, Pedro VV and Cogima, Camila K and Dezen-Kempter, Eloisa and Carvalho, Marco AG}, journal={Pattern Recognition Letters}, year={2020}, publisher={Elsevier} }
Most of the point clouds used to test our method are also availabe on CHAS dataset. Check it out, is public!
@dataset{paiva_pedro_victor_vieira_de_2019_2609498, author = {Paiva, Pedro Victor Vieira de and Cogima, Camila Kimi and Dezen-Kempter, Eloisa and Carvalho, Marco Antonio Garcia de}, title = {{CHAS - Cultural Heritage Architectural Segmentation dataset}}, month = mar, year = 2019, note = {{This dataset was funded by São Paulo Research Foundation trough e-Science program grants: #2016/04991-0, #2017/02787-9, #2017/01237-5.}}, publisher = {Zenodo}, version = 1, doi = {10.5281/zenodo.2609498}, url = {https://doi.org/10.5281/zenodo.2609498} }