ID-Card image quality assessment

This repository contains python code to assess the quality of ID Card images. It contains a preprocessing pipeline and various quality metrics.

For detailed information, please refer to the usage example.

This repository contains the code referenced by the Paper "Image Quality Assessment of Identity Cards Using Measures from Open Face Image Quality"

Installation

To create a virtual environment and install the required packages, run:

conda env create -f environment.yaml
conda activate oidiq

Usage

To run the quality assessment on a set of images, use the run.py script. You can specify the input CSV file containing image paths, output CSV file for results, and other parameters. For example:

python run.py --i data/test_files.csv --o data/results.csv --batch-size 16

Citation

@INPROCEEDINGS{11558141,
  author={Grote, Gregor and Tapia, Juan E. and Rathgeb, Christian},
  booktitle={2026 14th International Workshop on Biometrics and Forensics (IWBF)}, 
  title={Image Quality Assessment of Identity Cards Using Measures from Open Face Image Quality}, 
  year={2026},
  volume={},
  number={},
  pages={1-6},
  keywords={Printing;Modeling;Faces;Quality assessment;Image quality;Lighting;Biometrics;Algorithms;Pixel;Training;Biometrics;Explainability;OFIQ;Identity Cards;Artificial Intelligence;Quality Assessment},
  doi={10.1109/IWBF68042.2026.11558141}}
}

Disclaimer

This repository is provided for research purposes only. The code and materials are shared to support academic study and reproducibility.

The datasets used in this analysis are not distributed through this repository. They must be requested directly from their respective original sources and are subject to their individual licenses and usage restrictions.

If you have any questions regarding this repository or the research, please contact: gregor.grote@h-da.de.