Crowd Exemption Notifier

Abstract

Crowd management is a challenging problem in the pandemic situation to ensure public safety. To overcome the difficulties, many technologies have been incorporated to cope up with the problems. However, the effectiveness of these techniques is limited due to the density of the crowd changing from low to extremely high depending on the time. We propose a robust feature-based approach to deal with the problem of crowd management for people's security. We have evaluated our method using a dataset and have performed a detailed analysis. The crowd is monitored through a surveillance camera and the image is recognized frame by frame using OpenCV and the recognized frame is processed through the YoloV4 and COCO dataset to get the human identification count. When the crowd exceeds the allocated limit “Crowd Exemption Notifier” intimates the organizer. This mechanism could help the common people and the government to withstand the widespread viral disease.