REAL-TIME AI FRAMEWORK FOR DETECTING MISSING PERSONS IN CLOSED ENVIRONMENTS USING CCTV NETWORKS AND SPATIOTEMPORAL ANALYSIS

Abstract Due to the association of missing persons in inaccessible, as in visible closed environments such as retail businesses and public transportation, the search methods of today are inadequate. We propose a comprehensive, novel, real-time AI framework for the improved detection and tracking of missing persons using a newly developed spatiotemporal multi-camera integrated closed-circuit television (CCTV) with novel detection and re-identification of persons over video streams. This is achieved by combining cutting-edge computer vision and deep learning algorithms to develop robust features for detection and re-identification of persons through myriad streams. The framework continues to analyze movements in real time, perform predictive trajectories, and observe violations or abnormal behavior. The goal is to develop a highly sensitive, accurate, and automated system for tracking and locating missing persons rapidly to allow for more effective emergency responses. In general, we expect this framework to perform not only in terms of detection accuracy, but also in terms of tracking performance and time to commitment, which is superior to existing methods, and increase our overall capability to enhance public safety and surveillance, especially in confined indoor environments where standard systems fall short.