YOLO Based Smart Fall Detection Intergrated with Social Distancing Monitoring System

The research introduces a Smart Fall Detection and Social Distancing Monitoring System employing the YOLO (You Only Look Once) technique used for real time object identification and localization in a much faster way. The system processes video frames, converting them to black and white for efficiency, and uses YOLO to identify and classify individuals. Overall, the novel aspect of the research disagrees with the concept but focuses on the destination of the dimensions of the detected persons' bounding boxes. Once the height exceeds the width, an immediate alert follows regarding the alert of the detected fall. Therefore, the approach is non-invasive and occurs in real-time, which makes responses to possible emergencies as quick as. Those responsive measures are particularly beneficial for vulnerable categories of the population. In addition, the aspect of social distancing is accounted for by examining the spatial interaction between the individuals. Therefore, the adaptation of the tool is multiple at various locations and is consistent with the current health priorities. It positively affects safety and comfort. Thus, research indicates the possible implications of YOLO- based analysis of the video stream for the creation of real-time efficient solutions for health and wellbeing. The testing of the designed mechanism was performed on two publicly accessible datasets which has shown overall accuracy of 95% of this model, both as fall detection and social distance monitoring.