Spatially-Constrained Anomaly Detection in Crowded Environments using Meta-Heuristic Algorithm

Crowded-environment anomaly detection is crucial for public safety, surveillance, and crowd management. Traditional anomaly detection algorithms struggle with congested scenes. This study introduces a metaheuristic algorithm for spatially-constrained anomaly identification in congested environments. Metaheuristic optimization uses spatial limitations in the population to find anomalous patterns. The program can identify normal crowd behavior from aberrant by adding spatial linkages and boundaries between individuals. A metaheuristic algorithm intelligently searches for optimal constraint configurations to maximize normal-abnormal behavior discrimination. This adaptive optimization approach handles population densities, dynamics, and environmental factors. Proposed approach is tested on various crowded scene datasets. Proposed anomaly detection method outperforms others. Spatial limitations increase anomaly detection accuracy and robustness in overcrowded environments. This system has major applications. It detects violence, stampedes, and odd behavior in crowded settings to improve public safety. Eliminating false alerts and focusing on real anomalies can help crowd management and surveillance systems.