Inkjet Droplets Contours

This dataset provides a structured representation of droplet shapes generated in drop-on-demand (DoD) inkjet experiments, enabling quantitative analysis and data-driven modeling of droplet formation processes.

It consists of two main files:

1. HDF5 File: Contour-Based Representation

The HDF5 file contains the extracted contours of droplet boundaries, obtained from high-speed images available in the companion dataset (DOI: 10.5281/zenodo.13862494). Each group in the HDF5 file corresponds to a single experimental case (image). For each group:

1. • The input parameters (inkjet trigger signal characteristics) are stored as attributes.
   • The droplet contour(s) are stored as arrays of 2D points representing the boundary.
   • A parametrized representation of the contour is provided using Elliptic Fourier Descriptors (EFDs).

2. CSV File: Tabular Parametric Dataset

   The CSV file contains a tabulated dataset of 274,911 samples, where each row corresponds to one experimental case.  Columns include:

• • Trigger signal parameters (input variables)
  • Elliptic Fourier Descriptors (EFDs) representing droplet morphology

This file is particularly suited for statistical analysis, regression, and machine learning applications, providing a compact, low-dimensional representation of droplet shapes.

Data Processing and Tools:

• Image processing and contour extraction were performed using our open-source pipeline: https://github.com/DropletDynamics/InkJetDroplet
• Elliptic Fourier Descriptors were computed using: https://github.com/hbldh/pyefd

Licence:

This dataset is licensed under CC BY-NC-SA for non-commercial use.
If you modify or adapt the dataset, you must share the modified version under the same license (CC BY-NC-SA) and credit the original source.
For commercial use, a separate license is required—please contact droplets@cimne.upc.edu for details.