Computer vision as a tool for interpreting images

"Computer vision can save lives during sporting events," says Mubarak Shah of the University of Central Florida. Check out the video interview.

By José Tadeu Arantes

Agência FAPESP – Computer vision is a research area that brings together several disciplines and makes massive use of computing resources to obtain, process, analyze and interpret images, such as those obtained by a system of multiple cameras, to enable rapid and efficient decision making.

Developed primarily in the United States and associated with issues related to security and national defense, the scope of this field of study can be quite extensive in that it includes applications in civil defense, environmental monitoring and even biomedicine.

To present his outlook on the subject, Mubarak Shah, founding director of the Center for Research in Computer Vision at the University of Central Florida (UCF), was in Brazil to deliver a lecture at the Institute of Mathematics and Statistics at the University of São Paulo (USP).

Born in Pakistan but a long-time resident of the United States, where he has spent his career, Shah is the world’s leading authority on the subject and co-author of four books: Motion-Based Recognition (1997); Video Registration (2003); Automated Multi-Camera Surveillance: Algorithms and Practice (2008) and Modeling, Simulation and Visual Analysis of Crowds (2014).

Shah and his team have developed algorithms for purposes as varied as scanning crowds in search of suspicious people and conducting brain imaging analysis for identification of possible tumors.

“Computer vision can save lives,” the researcher told Agência FAPESP. “For example, at a sporting event with thousands of spectators, computer vision could help monitor the crowd so that in case of any commotion, people would not be blocked and get hurt or even killed. There is no limit to the use of computer vision for civilian purposes.”

Professor Shah’s lecture, held at the Institute of Mathematics and Statistics at the University of São Paulo (IME-USP), was organized by Roberto Marcondes Cesar Junior, professor in the Department of Computer Sciences at the IME-USP and member of the FAPESP Area Panel on Physics, Mathematics, Chemistry and Engineering. The event was part of the thematic project “Models and methods of e-Science for life and agricultural sciences.”

Another goal of Shah’s visit was to attract Brazilian students to the University of Central Florida, where they could enhance their training.

“Brazil today is one of the largest economies in the world, and we’re very interested in increasing our collaboration with the country. Our group at UCF is one of the world’s leading computer vision groups. We have 25 doctoral candidates and are hosting a large number of visitors from several countries. We would also like to host Brazilian students,” he said.