Perception Algorithms

A lightweight collection of perception algorithms published as supplementary material for research papers.

Overview

This repository contains reference implementations and experimental modules to support reproducibility and further research.

Running the Algorithms

You can run the algorithms in either MATLAB or Python.

Python

1. (Recommended) Create and activate a virtual environment.
2. Install dependencies:

pip install -r requirements.txt

3. Run a script from the repository root:

python classification/demo_dimension_based_classification_over_time.py
python classification/demo_dimension_based_classification.py
python dimension_estimation/demo_dimension_estimation_aeberhardt.py
python detection_and_tracking/demo_euclidean_clustering.py
python detection_and_tracking/demo_detection.py
python detection_and_tracking/demo_kalman_motion_model_over_time.py
python detection_and_tracking/demo_existence_probability_evolution.py
python detection_and_tracking/demo_kalman_detection_clustering.py

MATLAB

From the repository root in MATLAB, run:

addpath('classification', 'dimension_estimation', 'detection_and_tracking');

demo_dimension_based_classification_over_time
demo_dimension_based_classification
demo_dimension_estimation_aeberhardt
demo_euclidean_clustering
demo_detection
demo_kalman_motion_model_over_time
demo_existence_probability_evolution
demo_kalman_detection_clustering

Citation

If you use this repository, please cite the corresponding paper(s):

@article{schaefer2025lidar,
  author = {Sch\"afer, Simon and Alrifaee, Bassam and Hashemi, Ehsan},
  title = {Lidar-based Object Tracking of Traffic Participants with Sensor Nodes in Existing Urban Infrastructure},
  journal = {Advanced Intelligent Systems},
  volume = {n/a},
  number = {n/a},
  pages = {e202501102},
  doi = {https://doi.org/10.1002/aisy.202501102}
}