Paper: https://link.springer.com/article/10.1007/s11042-022-11923-2
Requirements
The repository has been tested in the following software.
- Ubuntu 16.04
- Python 3.7
- Anaconda
- Pycharm
Anaconda environment
To create and setup the Anaconda Envirmorent run the following terminal command from the repository folder:
$ conda env create -f env_MTMC.yaml
$ conda activate env_MTMC
Clone repository
$ git clone https://github.com/elun15/Online-MTMC-vehicle-tracking.git
Download AIC19 dataset
The dataset can be downloaded at https://www.aicitychallenge.org/track1-download/
Prepare AIC19 dataset
Move the downloaded folders aic19-track1-mtmc/train and aic19-track1-mtmc/test to the ./datasets/AIC19/ repository folder.
Preprocess the data to extract the images from the .avi files by running:
python preprocess_data.py
The set of data can be changed, by default it will preprocess /test/S02 scenario.
Download pretrained model
The model weights trained on AIC19 S01 scenario can be downloaded at: http://www-vpu.eps.uam.es/publications/Online-MTMC-Tracking/ResNet50_AIC20_VERI_layer5_imaugclassifier_latest.pth.tar
Place the weights file under ./models/
Training details can be found in the paper.
Running
To run the tracking algorithm over the S02 scenario run:
python main.py --ConfigPath ./config/config.yaml
If you find this code and work useful, please consider citing:
@article{luna2022online,
title={Online clustering-based multi-camera vehicle tracking in scenarios with overlapping FOVs},
author={Luna, Elena and SanMiguel, Juan C and Mart{\'\i}nez, Jos{\'e} M and Escudero-Vi{\~n}olo, Marcos},
journal={Multimedia Tools and Applications},
pages={1--21},
year={2022},
publisher={Springer}
}