Smart Surveillance System for Women and Children Using Machine Learning
Author : R Gomathi Jayam and S Suba Lakshmi
Abstract :
The safety of women and children is a major concern in modern society due to increasing incidents of harassment and violence. Conventional monitoring methods rely heavily on manual observation and passive video recording, often resulting in delayed responses and missed critical events. To address these limitations, this project proposes a Smart Surveillance System that leverages Machine Learning techniques for intelligent video analysis. The system is designed as a web-based application that allows users to upload video files for automated processing. It utilizes YOLOv8 (You Only Look Once Version 8) for real-time human detection and suspicious activity identification. When potential threats are detected, the system triggers an immediate visual alert through a full-screen flashing red overlay along with a siren sound. All incidents are automatically recorded in an SQLite database with timestamps for future reference and analysis. The proposed system emphasizes automation, rapid response, reliability, and ease of use. Experimental evaluation demonstrates high detection accuracy and prompt alert generation. This intelligent application-based solution provides a proactive safety mechanism suitable for homes, schools, hostels, and public environments.
Keywords :
Smart Surveillance, Machine Learning, YOLOv8, Real-Time Detection, Violence Alert, Women Safety, Children Safety, Automated Monitoring.
