DefinePK

Abstract

In a rapidly interconnected world, maintaining public safety is paramount. This work introduces a novel approach to utilizing CCTV camera feeds to identify real-time events like accidents and violent offenses. The system continuously analyzes video data by applying advanced computer vision methods and machine learning models, notably combining Long Short-Term Memory (LSTM) networks and Mobile Networks in Convolutional Neural Networks (CNN). Upon detection of an event, the system captures a short video record as evidence and reports it immediately through a chatbot, with ease of contact with the respective authorities. Through the chatbot, real-time alerts are generated, allowing swift action and softening the blow of such incidents. By synergizing automated event detection and real-time reporting, a solid mechanism for increasing public safety and creating a secure community is put into place