Intelligent Video Surveillance System: AI + IoT Solution for Safer and Smarter Urban Environment

Abstract

South Africa continues to grapple with persistently high crime rates that undermine public safety and socio-economic development. Conventional surveillance systems, Closed-Circuit Television- CCTV, rely heavily on human monitoring and post-event investigation, have proven inadequate in deterring violent crimes or ensuring timely response. This project explores the integration of Artificial Intelligence (AI) and Internet of Things (IoT) technologies to design an intelligent video surveillance system capable of real-time detection, anomaly analysis, and automated alert generation. By examining global smart city projects and solutions such as Smart Video Surveillance (SVS), Dubai Smart City, and ShotSpotter, this study identifies best practices and limitations relevant to the South African context. The proposed system emphasizes privacy-by-design, ensuring that only flagged incidents -such as weapon displays or violent altercations- are transmitted to control centres or local responders. This approach addresses concerns around privacy, governance, and misuse, while also providing strong evidentiary value for prosecutions in a justice system plagued by low conviction rates. Ultimately, this research highlights how AI-enabled IoT surveillance can contribute to building safer, more resilient cities by empowering communities, private security providers, and law enforcement agencies with timely, actionable intelligence.