Font Size: a A A

Research On Digital Video Monitoring System For Shaanxi Tianhong Silicon

Posted on:2015-10-21Degree:MasterType:Thesis
Country:ChinaCandidate:S YanFull Text:PDF
GTID:2298330467488625Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
Based on Tianhong Silicon Corp.’s practical need in association of practice, the thesisconstructs enterprise intelligent digital video surveillance system with the help of networktechnology, image processing technology and computer technology. The system is capable ofreal-time-detecting and analyzing to enterprises’ office areas, plant areas and warehouseswhile motion detecting and tracking, intelligent perimeter analyzing and fire-smoke detecting.Tianhong Silicon Corp.’s asset, staff and production security are ensured by improvingsecurity system and associating technical defense with human defense to achieve in modernmanagement improvement, energy-saving reduction, economy and social benefit promotion.The thesis focuses on the deep study of hardware in video surveillance system, afterintensive practical researches, spherical web-cams with PTZ and gun web-cams are used forvideo capture, and two-layer switchboards is used for network data transmission. To achieveintelligent perimeter surveillance and objects detection, moving object detection based onMixtures of Gaussians Modeling and target tracking algorithm based on Kalman filter areproposed. Based on color features, fuzzy and diffusion characteristics, a real-time algorithmis proposed for fire smoke monitoring, and fire alarm before the fire.Web-cam can run well after the system installation and adjustment, and it canautomatically detect and analyze Unknown vehicle and pedestrians. The system can beresponse in time when multi-users simultaneously query operations at the same time.The experiment show that the background modeling of Mixtures of Gaussians is close tothe actual video background, however the method is sensitive to noise like light change andcomputing complex. Objects tracking algorithm based on Kalman filter can confirm theoutline of target for intelligent perimeter analysis. Fire smoke detection algorithm issusceptible to noise interference, and it is easy to have a false alarm. This video surveillancesystem can be applied in production monitoring as well as security monitor in school andmall.
Keywords/Search Tags:video detection, web-cam, image processing, video analysis, target detection
PDF Full Text Request
Related items