Multi-Layered Edge Computing Architecture of Smart Information System for Cross-View People Tracking

Pedestrian detection and tracking have long been popular topics in both research and application. In intelligent building video surveillance systems, these techniques can be used to detect and track pedestrians within monitored areas, identifying suspicious behaviors. Currently, most widely-used pedestrian detection and tracking systems are based on a centralized architecture with a central server. Such systems suffer from high latency, poor scalability, and limited robustness. To address these issues, this paper proposes an end-to-end pedestrian detection and tracking system architecture utilizing edge computing devices. This architecture divides complex building structures into logical multi-layer structures using edge computing devices, solving the aforementioned problems, facilitating building security management, and reducing the complexity of pedestrian tracking algorithms in large buildings. Finally, a demo system is built to verify the feasibility of the proposed architecture.