Leveraging Smart Sensing for Proximity Analysis

This research delivers a deep learning-based approach to identify social distancing in interactions with others. The detection system is developed to analyze a video feed and also serves as a real-time detection tool to notify people to keep a safe distance from one another. This approach includes an alert system in voice form which alerts the people who are not following the protocols, provides the graphical insights to analyze the crowd density, and sends email alerts to higher officials to take control immediately. This application helps police officials monitor crowded areas during COVID-19 times. So, limiting human involvement in maintenance and enabling real-time auto-alert systems is the core concept of this research. It also provides data insights for the trend analysis in the form of a line graph which helps to make decisions quickly in critical situations. It identifies the areas that are overcrowded and allows police officials to take the necessary actions.