UruDendro: Cross-Section Images of Pinus Taeda

A dataset and Python toolkit for automated tree ring detection and analysis in Pinus Taeda cross-sections.

Paper • Dataset (Zenodo) • Project Page

---

Features

• 📊 Public dataset of tree ring images
• 🔍 Automated detection & segmentation
• 📏 Evaluation metrics
• 🎨 Annotation visualization
• 🤖 Deep learning background removal (U2-Net)

---

Quick Installation (Recommended)

Install uv if you don’t have it:

pip install uv

Create a fast, isolated environment and install all dependencies:

uv venv .venv
source .venv/bin/activate
uv pip install -r requirements.txt
pip install -e .

Note: The U2-Net model file (u2net.pth) must be in urudendro/. If you cloned with git-lfs:

sudo apt-get install git-lfs
git lfs pull

---

Usage

Download dataset:

import urudendro
urudendro.download_dataset('/absolute/path/to/dataset')

Visualize annotations:

import urudendro
urudendro.visualize_annotation('annotation.json', 'image.png', 'output_dir/')

Evaluate detection:

import urudendro
precision, recall, f_score, rmse, tp, fp, tn, fn = urudendro.compute_metrics(
    'detection.json', 'ground_truth.json', 'image.png',
    cx=512, cy=512, threshold=0.5, output_dir='results/'
)

Remove background:

import urudendro
urudendro.remove_salient_object('input.jpg', 'output.jpg')

---

Requirements

• Python ≥ 3.8
• PyTorch ≥ 2.4.1
• OpenCV ≥ 4.8.1
• NumPy ≥ 1.26.1
• See requirements.txt for full list

---

Citation

If you use UruDendro, please cite:

@article{marichal2025uruDendro,
  title={UruDendro: a public dataset of cross-section images of Pinus taeda},
  author={Marichal, Henry and Passarella, Diego and Lucas, Christine and Profumo, Ludmila and Casaravilla, Verónica and Rocha Galli, María Noel and Ambite, Serrana and Randall, Gregory},
  journal={Annals of Forest Science},
  volume={82}, number={1}, pages={1--21}, year={2025},
  publisher={Springer},
  doi={10.1186/s13595-024-01267-6},
  url={https://rdcu.be/euo3F}
}

Dataset: https://doi.org/10.5281/zenodo.15110646

---

Alternative Installation

You may also use pip or conda if preferred:

pip install git+https://github.com/hmarichal93/uruDendro.git
# or
conda env create -f environment.yml
conda activate uruDendro
pip install -e .