M. A. Hossen, P. E. Abas


With its massive and rapid spread, the ongoing COVID-19 coronavirus pandemic has caused devastation worldwide. Due to the lack of successful restorative medications and shortage of vaccinations against the virus, vulnerabilities among communities have surged. While a handful of countries has vaccinated most of their population, for many countries it is still a big challenge. Social distancing is considered to be an effective preventative measure against the spread of the pandemic virus, and virus propagation can be considerably curbed by limiting physical contact between people. The objective of this work is, therefore, to provide a depth image-based and cost-effective framework for social distance monitoring. A widely used human detection algorithm from depth image is used to estimate body joint position in real-time. To approximate social distance violations between people, the distance between individuals can be estimated, and compared to a predefined threshold. Outcomes of the work indicate that the proposed method can successfully identify individuals who violate social distancing rules, with over 98% detection accuracy. The result is significant, as it can be implemented in real-time to assist in the monitoring and enforcement of the new social distancing norm by all people, and thereby, curb the spread of COVID-19.

Full Text:



  • There are currently no refbacks.