This repository contains a python implementation of the Density Adaptive Point Set Registration (DARE) method. Without the density adaptation, the method is equivalent to Joint Registration of Multiple Point Sets (JRMPC) [1]. Additionally, implementations of Color-based Probabilistic Point Set Registration (CPPSR) [2] and Feature-based Probabilistic Point Set Registration (FPPSR) [3] are provided and can be run together with the density adaptation.
The script reg_demo.py runs DARE on a subsampled version of the vps outdoor dataset.
This method is also included in https://github.com/felja633/RLLReg as a pytorch implementation.
A detailed description of the DARE method can be found in the CVPR 2018 paper:
F. Järemo Lawin, M. Danelljan, F. S. Khan, P.-E. Forssen, and M. Felsberg, “Density adaptive point set registration,” in The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018. https://arxiv.org/pdf/1804.01495.pdf
@InProceedings{jaremo18a,
author = {Felix J\"aremo Lawin and Martin Danelljan and Fahad Khan and Per-Erik Forss\'en and Michael Felsberg},
title = {Density Adaptive Point Set Registration},
booktitle = {{IEEE} Conference on Computer Vision and Pattern Recognition},
year = {2018},
month = {June},
address = {Salt Lake City, Utah, USA},
publisher = {Computer Vision Foundation},
}
- python 3.6
- numpy
- scipy
- matplotlib
- pathlib
- cmake
- pcl (if you want to use FPPSR)
Make sure that the above dependencies are installed.
- To be able to run FPPSR, you need to build the pybind module in src/pcl_utils at src/pcl_utils/build.
You may use the shell script build_pybind_modules. The code has been tested in Ubuntu 16.04 and 18.04.
The full datasets used in the paper can be found at http://www.hdrv.org/vps/ and http://www.prs.igp.ethz.ch/research/completed_projects/automatic_registration_of_point_clouds.html.
Felix Järemo Lawin
email: felix.lawin@gmail.com
JRMPC: [1] G. D. Evangelidis, D. Kounades-Bastian, R. Horaud, and E. Z. Psarakis, “A generative model for the joint registration of multiple point sets,” in European Conference on Computer Vision, pp. 109–122, Springer, 2014 https://team.inria.fr/perception/research/jrmpc/
CPPSR: [2] M. Danelljan, G. Meneghetti, F. Shahbaz Khan, and M. Felsberg, “A prob- abilistic framework for color-based point set registration,” in CVPR, 2016. http://www.cvl.isy.liu.se/research/cogvis/colored-point-set-registration/index.html
FPPSR: [3] M. Danelljan, G. Meneghetti, F. Shahbaz Khan, and M. Felsberg, “Aligning the dissimilar: A probabilistic method for feature-based point set registration,” in ICPR, 2016. https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7899641