Study On Video Change Detection Service In Sensor Web

Video change detection is used in many areas including forest fire preventionand traffic management, etc. To meet these needs, a lot of video change detectionalgorithms and systems have been studied. The early technologies have a lot to learnfor us. However, with the development of information society, the demand of videochange detection is increasingly diverse. Thus, the existing system may face thefollowing challenges:1) A needed system is based on Web and standard interfaces;2)A needed system is real-time or near real-time;3) A needed system is on-demand.For this, a web system based on standard specifications is used to solve theseproblems. This system can support real-time task and is more flexible. On thefoundation of Web-based geoprocessing and Sensor Web technologies proposed byOpen Geospatial Consortium (OGC), we create a new method to construct suchsystems.Web Processing Service(WPS) defines Web-based standard serviceinterfaces for geoprocessing. Sensor Web defines standard Web service interfaces toshare and interoperate the data of sensor or sensor networks. The Sensor ObservationService (SOS), which is one of the OGC’s Sensor Web Enablement (SWE) services,manages sensor and sensor observation data using standard service interfaces.Firstly, the paper introduces research background and status at home and abroadof Sensor Web and video change detection, and significance of this study. Secondly,some key technologies about video change detection service in Sensor Web areemphasized: analysis of Sensor network service framework based on SWE, SensorObservation Service interfaces and Web Processing Service interfaces. Thirdly,primary metadata for video sensor modeling is analyzed in detail, which providesinformation model for video change detection. Then it systematically describessystem architecture of video change detection, including two sub-levelarchitectures—one is for real-time motion video change detection and the other is fordemand-based WPS. Finally, in order to confirm the function of the prototype system, some experiments on video observation and video processing services areexecuted. The results show that the proposed system architecture, model and keytechnologies for video change detection are feasible.