Smoke detection in video surveillance
Roberto Vezzani, Simone Calderara, Paolo Piccinini, Rita Cucchiara · 2008
ViSOR (Video Surveillance Online Repository) is a large video repository, designed for containing annotated video surveillance footages, comparing annotations, evaluating system performance, and performing retrieval tasks. The web interface allows video browse, query by annotated concepts or by keywords, compressed video preview, media download and upload. The repository contains metadata annotations, both manually created ground-truth data and automatically obtained outputs of particular systems. An example of application is the collection of videos and annotations for smoke detection, an important video surveillance task. In this paper we present the architecture of ViSOR, the build-in surveillance ontology which integrates many concepts, also coming from LSCOM, and MediaMill, the annotation tools and the visualization of results for performance evaluation. The annotation is obtained with an automatic smoke detection system, capable to detect people, moving objects, and smoke in real-time.