Hand detection from silhouette for video surveillance application

The new breakthrough in technology has opened new chances for researchers to develop more advance video surveillance system to provide human a more liveable city. Hand detection from silhouette for video surveillance application is important as it enables the tracking of human hand across video frames. This paper describes the algorithm for hand detection from human silhouette for video surveillance application. The Euclidean distance profile is calculated from the center point of silhouette to every pixel on the silhouette's edge profile. The first and the third significant peak of the profile were observed to denote the hand location in most cases (85%).