The July 2022 issue of IEEE Spectrum is here!

Close bar

Computer Count of Huge Crowds Now Possible

Automated crowd-counting software can reduce the time needed from up to a week to just half an hour

2 min read
Computer Count of Huge Crowds Now Possible
Demonstrators in Barcelona wave pro-independence Catalan flags during celebrations of Catalonia's National Day on 11 September. Software put the total number of demonstrators at 530,000.
Photo: Lluis Gene/AFP/Getty Images

Getting a headcount of crowds numbering in the hundreds of thousands need not strain human eyes any longer. New software has carried out the first automated crowd count on that scale ever by analyzing aerial photographs of a huge demonstrationa time-saving innovation that could eventually help save lives and prevent disasters.

The software, developed by University of Central Florida researchers, can drastically cut the time needed for crowd counts from a week to just half an hour. That represents the difference between the old method of humans painstakingly counting the number of heads per inch of aerial photograph and the new automated method. Such a speed-up in counting time could help event organizers and law enforcement decide how to safely handle huge crowds or act early to reduce the risk of stampedes and other dangers. 

“Automated computer analysis of such large-scale and dense crowds has never been done before,” said Mubarak Shah, computer science professor and director of the Center for Research in Computer Vision at the University of Central Florida, in a press release.

Shah and his colleagues tested their software on aerial photographs of a demonstration involving thousands of people calling for the independence of the Catalonia province from Spain. The software analyzed 67 aerial photographs taken of the demonstration in Barcelona and came up with a headcount within 30 minutes.

The images and software calculations were double checked by a team at the Pompeu Fabra University in Spain. Those researchers came up with a final headcount for the crowd of about 530,000—much smaller than the count claimed by the demonstration’s organizers.

Having automated crowd counts could make it much easier to verify attendance at huge outdoor events. Such crowd counts, especially for politically-charged events, have often been a source of contention between organizers and law enforcement.

The time-saving software could also make accurate crowd counts a crucial new tool for managing large crowds. Smart management of crowd movements could go a long way toward avoiding crowd disasters such as the Mina stampede during the annual Hajj pilgrimage that killed more than 1600 people in Saudi Arabia on 24 September, according to the latest death count from Associated Press. (For more, see the IEEE Spectrum article on how the Hajj crowds defy conventional computer simulations.)

Automated crowd counting could also prove increasingly potent with the growing use of aerial drone surveillance, whether in the hands of civilian authorities or militaries.

The Conversation (0)

The Future of Deep Learning Is Photonic

Computing with light could slash the energy needs of neural networks

10 min read
Image of a computer rendering.

This computer rendering depicts the pattern on a photonic chip that the author and his colleagues have devised for performing neural-network calculations using light.

Alexander Sludds

Think of the many tasks to which computers are being applied that in the not-so-distant past required human intuition. Computers routinely identify objects in images, transcribe speech, translate between languages, diagnose medical conditions, play complex games, and drive cars.

The technique that has empowered these stunning developments is called deep learning, a term that refers to mathematical models known as artificial neural networks. Deep learning is a subfield of machine learning, a branch of computer science based on fitting complex models to data.

Keep Reading ↓Show less