Bioframe

• 

  Landing page | Capture mode selection.

• 

  Upload of clear fingerprint image.

• 

  Final result of processed fingerprint image.

Inspiration

An accessible method of capturing fingerprints without the traditional live-scans but rather an object we already carry on daily basis, mobile phones.

What it does

This system captures fingerprints using a standard mobile phone camera making it ready for further biometric recognition and verification

How we built it

I built Bioframe solo using:

• OpenCV for the image capturing, ROI segmentation and preprocessing
  
• NumPy for efficient matrix (pixels) and feature computations
  
• Django REST Framework for the backend API
  
• A React application to handle the frontend and consume the API

Challenges we ran into

Extracting fingerprints from camera images that had various image quality and lighting conditions making it a bit difficult to develop a system that fits all.

Accomplishments that we're proud of

Successfully built a system for capturing fingerprints using software and camera and deploying on the web for easy accessibility

What we learned

How environmental factors could affect the quality of a fingerprint image such as light, skin tone and angle the fingerprint was captured from.

What's next for Bioframe

Matching algorithm to improve robustness to make this system industry ready for making capturing fingerprints easier for law enforcers, citizen registration etc.

Built With

• dbsqlite
• django
• numpy
• opencv
• python
• pythonanywhere
• react
• vercel