Tracking movement target by using WebGIS on video surveillance system

Abstract

This paper addresses a novel video surveillance system based on camera monitoring network. The system is belong to the application layer of Internet of Things (IoT). In the system, all videos are mapped into the corresponding camera nodes on a WebGIS (Web geographic information system). A group of the related nodes can be attained automatically by the Nearest Neighbor search method. In the searched node, the target can be automatically found in the video by target detection algorithm based on clothing color, and tracked based on particle filter. When above searching works are finished in an interesting geographical area, a completed movement path of the target can be displayed on the map, meanwhile, an information summary table about the target’s current location, current speed, current time and event’s attribute can be created. It also automatically collects the target’s relative video clips for improving the artificially monitoring work efficiency.