Research on the Estimation of Crowd Density Based on Video Image Processing

摘要

With the development of the technology and economic, the video supervising equipment are widely used in commercial, private areas and public management. It is innovative of the paper to make full use of the video images to estimation the crowd density and defend risks of stampedes. According to connect data in different crowds, this paper proposes a method that is based on the Gaussian Background Modeling and binary image to estimate crowd density. Results of this study provide a theoretical basis for the stampedes risk early warning.