INTRODUCTION
As the world welcomed 2020 , the city of Wuhan in China was fighting with newly discovered coronavirus “ COVID-19 .” Unfortunately , the timing of the emergence of this virus coincided with holiday season , with millions travelling around the world . The novel coronavirus turned out to be highly contagious and by early March , World Health Organization ( WHO ) declared the COVID-19 coronavirus as a pandemic 1 . Countries all around the world responded by imposing strict lockdowns restricting human interaction . While scientists are still working for development of immunization to this new virus , the world is trying to adjust to the new normal . A big part of COVID-19 response centers around safe return of employees and customers in public spaces such as retail shops , manufacturing plants , airports , cafeterias etc . This requires an ongoing need to monitor compliance with physical distancing and crowd density guidelines . To support business operations in these uncertain times , we propose a technology solution based on computer vision , streaming analytics and edge computing .
The objective of this project is to monitor physical distance compliance from video footage using deep learning and streaming analytics in a non-disruptive and anonymized manner targeted towards well defined spaces such as cafeterias , plant floors , retail checkout , airport security etc . Our solution surfaces insights via a real-time dashboard , an alerting mechanism , and a post-facto dashboard .
�
�
The real-time component can be deployed on-premise , configured to suit the target organization ’ s guidelines and provide immediate feedback via a real-time dashboard and push notifications to alert for urgent situations . The post-facto dashboard is meant to be deployed in a back-end environment such as the cloud and provides summarized feedback across multiple dimensions such as location , day of week and time of day . The dashboard supports interactivity to allow for what-if analysis and is a great tool to understand the efficacy of measures such as signage , oneway traffic etc . over a small period .
Few potential applications of the proposed solution are : �
Monitoring public spaces : Many public spaces already employ security cameras for surveillance . Utilizing the same setup , this solution can monitor physical distancing compliance in these public spaces .
1 https :// www . who . int / dg / speeches / detail / who-director-general-s-opening-remarks-at-the-media-briefing-oncovid-19---11-march-2020 .
- 2 - November 2020