Abstract

Nowadays, public safety is a key concern, particularly in fast urbanizing areas, urban development plans, and “smart city” projects where strong security measures have been implemented due to new security threats. The increasing trend of urbanization should be reflected in a coordinated approach to establishing efficient urban security systems. Current security systems have numerous constraints, including threat identification, on-the-spot analysis, and communication. These limitations become particularly emphasized in extremely complex urban areas with high populations and a broad range of activities. GuardianWatch, a groundbreaking AI-powered surveillance system, was developed in reaction to the above-mentioned difficulties with the primary goal of improving urban security. Real- time monitoring, detection, and alerting capabilities for a wide range of security concerns are precisely what GuardianWatch does. This includes identifying firearms, detecting physical aggressions, keeping an account of auto accidents, and identifying license plates. To achieve these goals under various urban settings and varying lighting and image quality issues, GuardianWatch employs a variety of cutting-edge artificial intelligence algorithms. This all-in-one innovative platform has tailored its algorithmic approach to the novel urban security settings with innovative artificial intelligence algorithms including YOLO v8, SK’s image models, TensorFlow, Haar cascades, and PyTorch.