• Kinjal Y. Joshi
• Safvan A. Vahora

Crowd behavior analysis is an active research topic in a video Surveillance for public security. Computer vision based approaches there are two types of behavior understanding problem: behaviors analyzer of individuals in a crowd (object- based) and using the knowledge to make deductions regarding the crowd behavior and analyze the crowd as a whole (holistic- based). In this proposed system is used holistic-based used, phase to analyze segmentation method using optical flow. Next we calculate clustering algorithm using k-means that dependence on the number of clusters (k), and the dependence on the initial choice of the clusters’ centroids. Moreover, our methodology guarantees creating high accurate clusters and defines Merging and Splitting of crowd behavior

Video Survillence, FAST Features, k-mean Clustering

"CROWD BEHAVIOR ANALYSIS USING FAST FEATURE", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.2, Issue 5, page no.1495-1500, May-2015, Available :http://www.jetir.org/papers/JETIR1505031.pdf

"CROWD BEHAVIOR ANALYSIS USING FAST FEATURE", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.2, Issue 5, page no. pp1495-1500, May-2015, Available at : http://www.jetir.org/papers/JETIR1505031.pdf