AI finger detection
A neural network finds and tracks every fingertip in the live preview – fast enough for real-time feedback on mid-range phones.
Touchless biometrics · Android & iOS
Biocapture turns the rear camera of a standard phone into an AFIS-ready fingerprint scanner. Hold up your hand – four fingers are detected, tracked and captured in one shot. No sensor, no contact, no extra hardware.
* 1000 dpi depending on camera hardware; 500 dpi on all supported devices.
How it works
Online or offline – the capture runs entirely on the device.
Scan your own hand – or, in second-person mode, an officer scans someone else's. The scan assistant guides placement; torch and exposure are controlled automatically.
A neural network tracks each fingertip while the hand moves from far to close and back, picks the sharpest frames, checks liveness and produces enhanced, calibrated fingerprint images at 500 or 1000 dpi.
Store on the device, send to a server for later processing, enroll into an AFIS database with identity data – or verify and identify against existing records.
Features
A neural network finds and tracks every fingertip in the live preview – fast enough for real-time feedback on mid-range phones.
An animated hand template shows where to place the fingers – for your own hand or, in second-person mode, for someone else's (e.g. police checks). Fewer retries, better first captures.
Photos, screens and printed fakes are rejected. Glare and parallax analysis runs passively while the user scans.
Each phone model is calibrated for true scale, so the output resolution is exact – easy enough for users to do themselves.
WSQ, JPEG 2000, PNG or raw bitmap; ANSI/INCITS 381, ISO/IEC 19794 and 39794, NIST Type 14 and Interpol XML.
Fingers that are missing or injured can be skipped. Single fingers, thumbs and four-finger slaps are supported.
Capture and processing happen on the device. No account needed; offline licensing available for disconnected field use.
Colors, images and texts of the scan screens are customizable, so Biocapture blends into your own app.
Security
Presentation attack detection (PAD) analyzes the live frames while the user scans – no extra step, no extra time. Both platforms share the same algorithms and thresholds.
Standards & AFIS
Tested against common fingerprint matchers and AFIS databases – and easily adapted to custom formats.
Supports matching against rolled digital samples and flatbed scans. * 1000 dpi depending on camera hardware.
Use cases
For developers
Drop the scan activity or view controller into your app, receive the finger images through a delegate. Documentation, sample app and release notes for each platform.
Plugins for Flutter, Cordova and Xamarin; React Native integration is possible. Other operating systems and custom APIs on request.
Try it
The Biocapture demo app shows the complete scan flow: capture, review and share your fingerprints in standard formats. No account, works offline. Captured fingerprints stay on your device unless you share them.
Allow installing from this source when your phone asks.
Install Apple's TestFlight app first, then open the invitation.
Licensing, evaluation keys, custom formats or a pilot with your AFIS – let's talk.
ms@ms-sw.at