A Review on AI-Powered Surveillance System using Computer Vision for Public Safety

Abstract

This review illuminates the rising role of AI-powered surveillance systems in enhancing public safety. Urban growth and evolving threats demand intelligent, automated monitoring solutions. Computer vision enables real-time video analytics, activity recognition, and object detection. These systems autonomously identify anomalies and aid law enforcement with alerts. Frameworks like Smart Watch and CV-Patrol use deep learning for rapid decision-making. Edge computing enhances performance across large- scale surveillance networks. Bots with multi-modal sensors offer 24/7 monitoring in public and high-risk areas. A layered approach—detection, verification, and response—boosts threat management. Advances include facial/vehicle recognition and behavior prediction models. Privacy-aware data handling ensures ethical surveillance deployment. Future goals involve fair AI models, better transparency, and agent coordination. This review outlines key innovations shaping AI-driven public safety systems