Survey of Crowd Flow and Density Estimation in Video Surveillance

With the development of economy and the increasing of crowd activities,crowd analysis algorithm emerges prominently due to the broad applications in surveillance.The crowd density estimation algorithm based on pixel statistics and texture analysis is investigated;and the pedestrians counting algorithm based on segmentation and statistics regression.The paper indicates both the strong point and defects of present algorithms.Improvement and promotions of the algorithms are needed in future research work.The developing trend of crowed analysis field is given in the last section.