Vehicles surveillance system provide a large range of informational services for the driver and administrator such as multi view road and driver surveillance videos from multiple cameras mounted on the vehicle, video shots highlighting the traffic conditions on the roads and monitoring driving behavior. Annotation and retrieving of these video streams has an importance role on visual aids for safety and vehicle guidance. Using video as a primary multimedia data source requires effective ways of retrieving the desired video data from a database. To do so, a model that classifies vehicle video data on the basis of traffic information and its semantic properties which were described by driver's eye gaze direction was developed in this paper. Then the annotated video data based on the model is organized and streamed by retrieval platform and adaptive streaming method. The experimental results show that this model is a good example for evidence-based traffic instruction programs.