🪵 IPAN 3D

3-D Granulometry Using Image Processing

MATLAB source code for the IEEE Transactions on Industrial Informatics paper
3-D Granulometry Using Image Processing

---

✨ Overview

IPAN 3D is a MATLAB implementation for estimating the three-dimensional granulometry of particles using image processing techniques.
The project targets industrial scenarios where the size distribution, shape, and geometry of particles or material fragments must be measured automatically from images.

The method was presented in the IEEE TII 2019 paper:

R. Donida Labati, A. Genovese, E. Muñoz, V. Piuri, and F. Scotti,
“3-D Granulometry Using Image Processing,”
IEEE Transactions on Industrial Informatics, vol. 15, no. 3, pp. 1251–1264, March 2019.

---

🧭 Processing Workflow

flowchart LR
    A[📷 Stereo image acquisition] --> B[🎯 Calibration data]
    B --> C[🧩 Image preprocessing]
    C --> D[📐 3-D reconstruction]
    D --> E[🪵 Particle size estimation]
    E --> F[📊 Granulometric measurements]

Loading

The pipeline combines stereo captures and calibration information to estimate particle dimensions and produce granulometric measurements suitable for industrial analysis.

---

🖼️ Visual Examples

---

📁 Repository Structure

IPAN_3D/
│
├── Code/                         # MATLAB source code
│   └── launch_IPAN_caduta_size    # Main entry point
│
├── Dati_recenti_DEMO/             # Demo database of recent stereo captures
├── Dati_storici_DEMO/             # Demo database of historical data
├── calib_ren/                     # Calibration resources
│
├── fileParametriTutti.txt         # Parameter file
├── numcams.txt                    # Camera configuration
├── tempcode.txt                   # Auxiliary configuration/code file
│
├── README.md
└── LICENSE

---

🚀 Getting Started

1. Clone the repository

git clone https://github.com/AngeloUNIMI/IPAN_3D.git
cd IPAN_3D

2. Open MATLAB

Start MATLAB and move to the repository folder:

cd('path/to/IPAN_3D')

3. Check the demo data

The repository includes demo stereo-capture data in:

Dati_recenti_DEMO/
Dati_storici_DEMO/

The current README identifies Dati_recenti_DEMO as the required database of stereo captures for the main demo.

4. Run the main script

From MATLAB, launch:

run('./Code/launch_IPAN_caduta_size')

Depending on your MATLAB path configuration, you may need to add the source tree first:

addpath(genpath('./Code'))
run('./Code/launch_IPAN_caduta_size')

---

📊 What the Code Does

IPAN 3D is designed for automatic granulometric analysis from image data.

Stage	Purpose
Stereo acquisition	Uses paired views of falling or moving particles
Calibration	Uses camera calibration information to support 3-D reasoning
Image analysis	Extracts visual information from captured particle images
3-D estimation	Estimates particle geometry from multi-view data
Granulometry	Produces size-related measurements for industrial analysis

---

🔗 Related Dataset: Granulo-10k

For newer work on industrial granulometry, see the related repository:

Granulo-10k — A large-scale benchmark dataset for multiple-view industrial granulometry of OSB wood strands

Granulo-10k provides a modern benchmark with:

• 9,600 RGB images at 1280 × 960
• 200 OSB wood strands
• 24 acquisitions per strand
• 2 synchronized camera views per acquisition
• segmentation masks
• 3-D point clouds
• ground-truth height, width, and thickness measurements

---

📖 Paper

If you use this repository, please cite:

@Article{ipan3d,
  author  = {R. {Donida Labati} and A. Genovese and E. Muñoz and V. Piuri and F. Scotti},
  title   = {3-D granulometry using image processing},
  journal = {IEEE Transactions on Industrial Informatics},
  volume  = {15},
  number  = {3},
  pages   = {1251--1264},
  month   = {March},
  year    = {2019},
  note    = {1551-3203}
}

Paper: IEEE Xplore
Project page: IPAN 3D Project

---

👥 Authors

• Ruggero Donida Labati
• Angelo Genovese
• Enrique Muñoz
• Vincenzo Piuri
• Fabio Scotti

Department of Computer Science
Università degli Studi di Milano, Italy

---

📄 License

This project is released under the GNU General Public License v3.0.

See the LICENSE file for details.

---