Particle Tracking for Soft Matter

Particle Tracking for Soft Matter offers a modular set of Jupyter notebooks to help experimentalists and researchers analyze microscopy data with minimal coding effort. The tutorials are designed to be accessible, with well-structured utility modules and clear workflows.

---

Overview

This repository supports the full pipeline for particle tracking:

• Detection – Identify and localize particles using both classical and deep learning methods.
• Linking – Associate particles across time frames to reconstruct trajectories.
• Simulation - Generate realistic dataset for training and evaluation.
• Evaluation – Compare predictions to ground truth using tracking metrics.
• Visualization – Animate and inspect results interactively.

---

Tutorials

The tutorials are organized into two main parts: Detection & Localization and Linking.

Detection & Localization

You will apply and compare several detection strategies:

1. Thresholding & Connected Components – Simple and fast.
2. Crocker–Grier (TrackPy) – Classical approach to particle tracking.
3. U-Net – Supervised deep learning for segmentation.
4. LodeSTAR – Unsupervised deep learning for subpixel localization.

Each method is benchmarked on simulated data and then applied to experimental datasets.

Tutorials:

• Detection of Spheres
  

• Detection of Core-Shell Spheres
  

• Detection of Ellipses
  

---

Linking

Associate localized particles across frames to reconstruct trajectories using:

1. Nearest-neighbor linking (TrackPy).
2. Linear Assignment Problem (LAP) using Hungarian algorithm (LapTrack).
3. Graph-based deep learning linker (MAGIK, via deeplay).

Tutorial:

• Linking Spheres
  

---

Getting Started

1. Clone the repository:

git clone https://github.com/softmatterlab/ParticleTracking.git
cd ParticleTracking

2. Install dependencies:

pip install -r requirements.txt

3. Launch the tutorials:

jupyter lab  # or jupyter notebook

---

Using utils/ in Google Colab

The tutorials rely on the utility modules stored in the utils/ folder.

Option 1: Clone the full repository (recommended)

!git clone https://github.com/softmatterlab/ParticleTracking.git
%cd ParticleTracking

Option 2: Upload the utils/ folder manually

Upload a zipped copy of utils/:

from google.colab import files
uploaded = files.upload()  # Upload utils.zip

Unzip it:

!unzip utils.zip -d .

Then import as usual:

from utils import detection_utils, tracking_utils, video_utils

---

Dependencies

Core libraries:

• numpy, scipy, matplotlib
• scikit-image, torch
• trackpy, laptrack, deeptrack, deeplay

See requirements.txt for full details.

---

License

This project is licensed under the MIT License.

---

Citation

If you use this toolkit for your research, please cite:
(BibTeX and citation information coming soon)