Crowd Counting App

Authors

• Pratyush Kumar Kar
• Aditya Sadangi
• Siddhanta Choudhury

Problem Statement

To develop an Event Management App that leverages the power of MobileNetV2 in OpenVINO for real-time object detection to enhance event planning and execution. This app should enable event organizers to efficiently monitor and manage crowd counting ,crowd dynamics, ensuring optimal safety, resource allocation, and participant experience during various events, from concerts and conferences to sports matches and festivals. The app should provide a user-friendly interface, allowing organizers to access live video feeds, receive crowd insights, and make informed decisions to improve event outcomes.

Assumption

1. Camera Placement : Either 4-5m from the ground , or headlevel of the crowd if necessary. (drone view preferred )
2. Visibility : Atleast 10-15 percent of the person’s body should be visible to register as a person , from any height or angle above eye level
3. Privacy and Ethical Considerations : Since no personal data on any detected person is stored or used , and people are rather used like objects in this model , it does not require many privacy based agreements to run in public places.