AIoT-Based Public Vandalism Detection and Alert System: A Review

This research paper presents an AIoT-Based Public Vandalism Detection and Alert System that aims to identify graffiti and spray-writing activities using artificial intelligence and Internet of Things (IoT) technologies. The proposed system integrates webcam-based video monitoring, OpenCV image processing and a YOLO-based detection model to identify vandalism in real time. When graffiti is detected above a predefined confidence threshold, Python communicates with an Arduino board through serial communication to activate an LED and buzzer. A Streamlit dashboard provides detection updates, including confidence scores and timestamps. The paper discusses the system architecture, methodology, existing research, applications, advantages and limitations. The developed prototype demonstrates how AI and IoT can be combined to support automated public infrastructure monitoring and improve the detection and reporting of vandalism.