Nowadays, visual surveillance systems populate many different corners in our cities. However, and contrary to what is popular belief, human identification in surveillance context remains an open problem so far.

Researchers from IT Covilhã and IT Lisbon, working in project QUIS-CAMPI, are developing a system to automatically identify persons in crowded environments, able to work in fully covert conditions, inside a delimited space, under conditions similar to the currently deployed visual surveillance systems. There are two main steps of the QUIS-CAMPI processing chain. The first step analyses the whole scene and, through the use of PTZ devices (cameras with remote directional and zoom control), it enables extracting high resolution images of the subjects heads and silhouettes of the walking person. The second step performs biometric recognition, by analyzing the subjects´ physical and behavioral characteristics.

By bridging the gap between biometric recognition and visual surveillance systems, QUIS-CAMPI represents a step towards the future of surveillance.