Elegant and efficient algorithms for real time object detection, counting and classification for video surveillance applications from single fixed camera

Video Surveillance is very important and essential task for security applications. Earlier surveillance was like capturing a video from camera, storing the information in a database, and then required contents were accessed manually from the database. It may lead to loss of sensitive information in real time. In such cases automated video surveillance is very essential. In automated video surveillance, object detection and tracking can be done in real time, and it finds the required information, also informs to the administrator in real time. This paper describes the detection of objects in real time and counts the number of objects. It also describes the objects classification; it is classified in to five predefined classes namely human beings, cars, motor bikes, busses and horses by the method of features extraction and Comparision.