Development of a Large-Scale Edge Image Detection System for Intelligent Urban Disaster and Safety Control

Recently, efforts have been increasing to effectively respond to disasters and safety issues that are increasing globally through intelligent city control using AI technology and CCTV. However, there is a problem in utilizing high-cost AI service infrastructure such as many AI-embedded CCTVs or high-level cloud computing resources for large-scale city control. This paper proposes the structure of an edge image detection system and an AI processing method that provides optimal scalability and resource efficiency through distributed processing and efficient distribution of resources in an intelligent edge system. Through this, it is possible to flexibly support the application of various edge-level models while providing the economic feasibility of expanding AI-based infrastructure for future city control.