How We Built PrezScan: Free Hardware-Free Biometric Attendance for Indian Universities

Описание и применение

PrezScan – бесплатное веб‑приложение для биометрического учета посещаемости в индийских университетах, использующее распознавание лица и геофенсинг Wi‑Fi, без аппаратного обеспечения и с гарантией конфиденциальности.

Задача

Hundreds of students line up daily before faulty fingerprint machines, causing delays.

Результат

PrezScan eliminates hardware failures, reduces queue times, ensures privacy by keeping raw photos local, and provides instant audited attendance reports.

Рабочий процесс

• Students open the PrezScan web app on a device
• The app captures a 68‑point facial image via the camera
• Client‑side WebGL GPU processes the image to a 128‑dimensional vector
• The app checks the student’s Wi‑Fi subnet to confirm campus location
• If both checks pass, attendance is recorded locally and synced to the server
• Admins export attendance as audited Excel reports with timestamps and session hours

Доказательства

Every morning across engineering colleges and universities in India, hundreds of people line up in front of wall-mounted fingerprint machines that break down and cause morning delays.

Источник

Dual‑Lock Anti‑Proxy Attendance: Enforces a physical double‑lock combining on‑device 68‑point AI facial recognition with academic building Wi‑Fi subnet (CIDR) geofencing.

Источник

Sub‑200ms Verification: Runs client‑side directly in the browser engine using WebGL GPU acceleration.

Источник

100% Privacy by Design (DPDP Act 2023 Compliant): Raw face photos are never sent over the internet or stored on servers; only local 128‑dimensional mathematical vector embeddings are evaluated.

Источник

1‑Click Audited Excel Reports: Monthly attendance ledgers export complete timestamps and session hours instantly in formatted spreadsheets.

Источник