Learning and Classification of Suspicious Events for Advanced Visual-Based Surveillance

The recent evolution of advanced visual-based surveillance (AVS) systems has allowed to introduce digital image processing and computer vision techniques in several application domains where a human operator has to observe multiple images provided by complex remote environments. The main goal of an AVS system is to generate automatically focus-of-attention messages in order to help the human operator to concentrate his decision capabilities on possible danger situations. In this way, possible human failures are expected to be overcome and better surveillance performances should be obtained [1].