Skip to content

irhdab/open-iris

 
 

Repository files navigation


IRIS: Iris Recognition Inference System

python Ruff isort pydocstyle
ci-testing release-version


the issue

Table of Contents

About

IRIS is Worldcoin origin open-source Iris Recognition Inference System, designed for robust, scalable, and secure biometric identification. This fork of it now uses a hash-based matching approach, generating a 40-bit unique identifier for each iris, enabling O(1) lookup and cryptographic security at global scale.

Features

  • Hash-Based Matching: Each iris template is converted to a 40-bit unique ID using SHA-256, enabling exact, fast, and secure matching.
  • O(1) Lookup: Matching is performed by comparing 40-bit integers, making the system highly scalable for billions of users.
  • Cryptographic Security: Uses SHA-256 for template hashing, ensuring strong privacy and resistance to tampering.
  • Modular Pipeline: Includes segmentation, feature extraction, and template generation.

Installation

git clone https://github.com/irhdab/open-iris.git

Performance & Security

  • Storage: Each iris is represented by a 40-bit (5 byte) unique ID, reducing storage by over 1,000x compared to traditional templates.
  • Speed: Matching is O(1) (integer comparison), suitable for massive-scale deployments.
  • Security: SHA-256 hashing ensures cryptographic strength and privacy.

License

This project is licensed under the MIT license.


About

Open Iris Recognition Inference System (IRIS)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 88.2%
  • Jupyter Notebook 11.7%
  • Makefile 0.1%