Surveillance and inspection are core topics in most automation applications where objective and/or manual monitoring is not realistic. Surveillance normally refers to monitoring animal and people, where inspection mostly refers to monitoring objects in a production setup, e.g., on a conveyer belt. Typically, sensors capture data, which are then processed using AI algorithms. The captured data and the processed results can be highly sensitive and security is therefore a third key component to ensure the integrity of the applications.
At AAU we have a long tradition of research in AI within surveillance and inspection using especially video data - ranging from detecting abnormal behavior among pedestrians to classical machine vision research aimed at robotic solutions to handle man-made objects on production lines.