Feature based approach for Efficient Crowd Density Estimation

Presently there is major interest in visual surveillance systems for crowd analysis. For preventing the problems caused by large crowd, proper control and management is necessary. In this paper a novel individual feature is introduced and used in combination with the holistic features, to describe crowd density. Various databases are analyzed to check the validity of the feature extraction process. In order to cluster the crowd frames according to congestion degree Support Vector Machine and Artificial Neural Network are used as classifiers. Results obtained shows that the proposed feature performs well in classifying real-world crowd scenes.