It is a daunting task to get a headcount of hundreds of thousands of people in a crowd. However, new software has been developed that can carry out an automated crowd count – the first of its kind. The groundbreaking software analyzes aerial photographs of such enormous crowds in tens of minutes – an innovation that could prevent disasters and save lives.
The software was developed by researchers at the University of Central Florida. It can reduce the time needed to count crowds from over a week to just 30 minutes. This speed in headcounting can help law enforcement and event organizers to safely deal with enormous crowds by acting early if something happens and reducing the risk of crowd-related disasters, such as stampeding.
Mubarak Shah, a computer science professor and director of the Center for Research in Computer Vision at the University of Central Florida stated in a press release that, “Automated computer analysis of such large-scale and dense crowds has never been done before.”
Shah and his team tested their invention on aerial photographs taken of a demonstration in Spain that involved hundreds of thousands of people. The software analyzed almost 70 photographs taken of the crowd and within 30 minutes, came up with a headcount.
The photographs and software calculations were then re-checked by a research team at the Spain’s Pompeu Fabra University. The researchers came up with a crowd headcount of about 530,000—much smaller than the count claimed by the demonstration organizers.
Having headcounts determined with such quickness could make it much simpler to verify the number of people in attendance at large outdoor events. Such headcounts have often created tension between law enforcement and event organizers, who often have a different estimate of the number of people present.
The software could also make accurate crowd counts a critical tool for crowd management. The overseeing of accurate crowd numbers could help avoid crowd disasters such as the Mina stampede that took place during the annual Hajj pilgrimage that killed more than 1,600 people in Saudi Arabia in late September.
Automated, time-saving crowd counting could also prove very useful for drones operated by both civilian and military authorities.