Distributed smart camera network for safety and security
Abstract
Current CCTV surveillance solutions are generally retrospective tools. Because real time use of CCTV requires human monitors to view a potentially exorbitant number of video feeds, CCTV is usually only useful after an incident has occurred. However, new technologies are making it possible for machines to perform some tasks that previously required a human monitor. The proposed project seeks to augment existing CCTV systems with behavioral analytics. The system uses a series of cameras, FPGAs, and computers to track object movement throughout a facility. This information is used to build a model of normal movement. Object movements are compared against this model and any ones that diverge from the model are flagged for review by security personnel.