Facial recognition system for crowd security

Giam, Tia-kaztenie Hui Zhi (2023) Facial recognition system for crowd security. Final Year Project, UTAR.

Preview

PDF Download (10Mb) | Preview

Abstract

Public safety is a top priority, but crowded areas witness numerous crimes annually, posing a threat to global peace and security. Identifying criminals and potential threats before they commit heinous acts like bombings, mass shootings, child abduction, and sexual assaults in public spaces is vital. While CCTV cameras offer post-incident monitoring, integrating facial recognition technology with live video feeds can proactively prevent such tragedies. A facial recognition system is developed to identify known criminals and missing persons from a face database, enabling public surveillance cameras to track their whereabouts, monitor their activities, and notify authorities promptly when needed. This Python project uses Convolutional Neural Network (CNN) face recognition with Dlib and Haar Cascade Classifier to effectively detect and monitor known and potentially dangerous individuals in publics areas, facilitating swift emergency responses when necessary, while keeping watch for missing persons. The system developed uses Firebase’s Realtime Database and Storage Bucket to store and retrieve data in real-time to expedite system functionalities like reports generation and database management.

Actions (login required)