Real Time Social Distancing Detection

Social distancing has become essential to prevent the spread of infectious diseases, especially in crowded areas. This paper proposes a real-time social distancing detection system using deep learning, specifically YOLOv3 (You Only Look Once version 3) object detection algorithm. The proposed system uses surveillance cameras to capture real-time video feeds of crowded areas, which are then processed by YOLOv3 to detect individuals and calculate their distances from each other. The system employs a custom object detection model trained on a large dataset of annotated images and a novel distance measurement method, which takes into account the variations in camera perspectives and distances of individuals from the camera. The experimental results demonstrate its high accuracy and real-time performance. The proposed system can contribute to the development of smart cities and public health systems that helps researchers and practitioners to design and implement real-time social distancing monitoring systems for various applications.