EFFICIENT SCALE INVARIENT FEATURE BASED METHOD FOR CROWD LOCALIZATION

Visual surveillance has been a very active research topic in the last few decade due to growing importance for security in the public areas. With the increasing number of CCTV networks in public areas, the enhancement in the computing power of modern computers and increase the possibility to entrust an automatic system with the security and the monitoring of events involving large crowds is within reach. Crowd detection and localization in the surveillance video is the first step in automatic crowd monitoring system. The performance of the whole system depends on this step. Detecting the crowd is a challenging task because the crowds come in different shape, size and color, against cluttered background and varying illumination conditions. As the size of the crowd increases managing the crowd becomes more complex