track-sort-py 0.3.1

High Performance implementation of the SORT object tracking algorithm

Track SORT

This is a implementation of the SIMPLE ONLINE AND REALTIME TRACKING paper [1], or SORT for short, for Python.

It essentially wraps the Rust implementation from https://gitlab.com/leandrosansilva/track-sort-rs

Its interface differs considerably from the original reference implementation from https://github.com/abewley/sort.

Installation

Requirements

• A recent version of Cargo, the Rust package manager.
• Maturin

Via PIP

pip install track-sort-py

TODO

• Document it
• Benchmark it against reference implementation
• Publish linux binaries

Citation

@inproceedings{Bewley2016_sort,
  author={Bewley, Alex and Ge, Zongyuan and Ott, Lionel and Ramos, Fabio and Upcroft, Ben},
  booktitle={2016 IEEE International Conference on Image Processing (ICIP)},
  title={Simple online and realtime tracking},
  year={2016},
  pages={3464-3468},
  keywords={Benchmark testing;Complexity theory;Detectors;Kalman filters;Target tracking;Visualization;Computer Vision;Data Association;Detection;Multiple Object Tracking},
  doi={10.1109/ICIP.2016.7533003}
}

[1] https://arxiv.org/pdf/1602.00763