The experimental code of 3DCDNet on Urb3DCD dataset
python 3.7.4
torch 1.8.10
visdom 0.1.8.9
torchvision 0.9.0
The Urb3DCD Dataset can be found from ( I. de Gélis, S. Lefèvre, and T. Corpetti, “Change Detection in Urban Point Clouds: An Experimental Comparison with Simulated 3D Datasets,” Remote Sensing, vol. 13, no. 13, p. 2629, Jul. 2021, doi: 10.3390/rs13132629.)
The pretrained model fro Urb3DCD is available at [Google Drive] and [Baiduyun] (the password is: txx5).
Before test, please download datasets and pretrained models. Change path to your data path in configs.py. Copy pretrained models to folder './outputs_{sub_dataset_names}/best_weights', and run the following command:
cd 3DCDNet_ROOT
python test.py
Before training, please download datasets and revise dataset path in configs.py to your path.
cd 3DCDNet_ROOT
python -m visdom.server
python train.py
To display training processing, open 'http://localhost:8097' in your browser.
The experiments on SLPCCD dataset can be foud from this link.
If you use this repository or would like to refer the paper, please use the following BibTex entry.
@ARTICLE{10184135,
author={Wang, Zhixue and Zhang, Yu and Luo, Lin and Yang, Kai and Xie, Liming},
journal={IEEE Transactions on Geoscience and Remote Sensing},
title={An End-to-End Point-Based Method and a New Dataset for Street-Level Point Cloud Change Detection},
year={2023},
volume={61},
number={},
pages={1-15},
doi={10.1109/TGRS.2023.3295386}}