Edge Computing-based Real-Time Surveillance System with YOLOv8 Object Detection using NVIDIA Jetson Nano

Conventional surveillance systems depend on centralized cloud architectures, resulting in high latency, excessive bandwidth usage, and potential security risks. These drawbacks limit real-time threat detection and delay response times, making them less effective for modern smart city applications. To overcome these challenges, this study presents an edge-powered smart surveillance system integrating a Jetson Nano kit and camera sensors for localized data processing. By conducting real-time video analysis at the edge, the system minimizes network congestion, improves response efficiency, and strengthens data privacy and security. Meanwhile, cloud infrastructure is utilized for long-term storage and advanced analytics, forming a hybrid approach that optimizes resource allocation. Initial evaluations indicate that this method surpasses traditional surveillance models in terms of speed, operational efficiency, and reliability. This research highlights the potential of combining edge computing with cloud-based systems to create a scalable and intelligent surveillance solution for smart cities.