Human Detection from Video Surveillance Using Silhouette - Based Method

Video surveillance is rapidly increasing meaningful approach for deterrent to crime, and ability to apprehend a suspect when a crime occurs. Before few years video surveillance paired with security guards which means security guards should watch surveillance TV at 24/7 days it made more pricy. But nowadays it’s turned into automatic which gives good safety with low budget. Normally Surveillance system is unification of three phases Object segmentation, Object classification, Object Tracking and Our proposed system design is directed on automatic classification of objects in video surveillance. In this paper, mainly we magazine previous research on object classification and analyze experimental results, finally provide our conclusions for improved performances. One stationary camera has been used for retrieve input. Keywords— Video surveillance, background subtraction, contour formation, object classification.