An Intelligence Traffic Accidence Monitor System Using Sensor Web Enablement

it is an urgent task to establish freeway network monitoring system. The capability of fast detecting traffic incidents and abnormity is necessary to reduce the loss caused by traffic accidents, to carry into accident rescue timely and effective, to reduce of traffic delay generated by incidents and to avoid the second accident. Using image recognition technology to automatic detect traffic incident via the existing highway surveillance cameras, is a very promise method. However, the existing video based traffic monitor systems do not interoperation and share mechanism to aggregate and cooperate the information as a whole. Hence, this paper proposes a new architecture through Sensor Web Enablement (SWE) interoperability interfaces and metadata encodings that enable real time integration of heterogeneous sensor webs into the information infrastructure, where traffic sensors can be Web accessible, traffic camera and its data can be discoverable, and traffic camera will be self-describing to humans and software (using a standard encoding). All traffic camera observations will be easily accessible in real time over the web.