Biometric Access Control

Overview

Biometric Access Control is a fingerprint recognition system built using OpenCV and machine learning techniques. It compares input fingerprints against a stored database and determines a match based on keypoint descriptors.

This project is based on the old research done by:

• Kristijan Jankoski
• Gabriel Oliveira
• Sashwat K
• Utkarsh Deshmukh

Additionally, deprecated commands and syntax errors from the original model (python-fingerprint-recognition) have been fixed to ensure compatibility with modern Python and OpenCV versions.

Features

• Fingerprint preprocessing and enhancement
• Feature extraction using Harris corner detection and ORB descriptors
• Matching system using the BFMatcher algorithm
• Visualization of matching fingerprints
• Automated comparison with a stored fingerprint database

Installation

Prerequisites

Ensure you have the following installed:

• Python 3.x
• OpenCV
• NumPy
• Matplotlib
• scikit-image

Setup

Clone the repository:

git clone https://github.com/cybereagle2001/Biometric_Access_Control.git
cd Biometric_Access_Control/recognition_model

Install required dependencies:

pip install -r requirements.txt

Usage

To run the fingerprint recognition program, execute the search.py script:

python3 search.py <fingerprint_image>

• <fingerprint_image>: Path to the fingerprint image you want to check.
• The script will compare the input fingerprint with the database and display the match result.

Project Structure

Biometric_Access_Control/
│── recognition_model/
│   ├── app.py              # Main fingerprint recognition script
│   ├── enhance/            # Image enhancement functions
│   │── database/           # Fingerprint dataset
|   │── search.py           # Main entry point for fingerprint search
│── requirements.txt        # List of dependencies
│── README.md               # Project documentation

Acknowledgments

Special thanks to the original contributors whose research and projects helped shape this implementation.

License

This project is open-source and available under the MIT License.