Regional Crowd Status Analysis based on GeoVideo and Multimedia Data Collaboration

Abstract

Multiple scene video surveillance systems can be combined through GIS technology to eliminate scene restriction and complete regional crowd state analysis. Regional surveillance video data is huge and complex, and its data screening, processing and analysis are difficult. Aiming at the problem of regional crowd state monitoring, this paper proposes a method of crowd state analysis based on GeoVideo and multimedia data cooperation. This method integrates social media data information, uses urban POI data and microblog sign-in data to extract hotspot areas to screen or collect the monitoring video of the hot area, then extracts the crowd state feature of the video and combines the VideoGIS to complete crowd status analysis in this area. Finally, the experiments which were designed for crowd flow monitoring confirm the feasibility and correctness of the method on the regional crowd state monitoring and analysis.