A Number Of Key Technologies Of Intelligent Video Surveillance System

Nowadays, video surveillance system is playing an important role in all ways of our life. On the whole, the development of video surveillance systems has experienced four stages: analog video surveillance systems, PC-based video surveillance systems, embedded system-based video surveillance systems and intelligent vision-based video surveillance systems.There are six key techniques in intelligent vision-based video surveillance systems. They are compression of digital video data, net transmit of digital video data, store and index of digital video data, receiving and playing of digital video data, motion detection, and object detection and recognition.This paper focuses on the research and implementation of these six key techniques. Firstly, this paper researches the international standards for compression of digital video data, analyses how to compress analog video stream to MPEG-4 with hardware. It is also involved in the techniques of net transmit of digital video data, such as net protocol, IP multicast, RTP/RTCP and so on.Secondly, in this paper, a new method in store and index of digital video data is studied. Additionally, DirectShow is used in the receiving and playing of digital video data.Thirdly, intelligent vision is so important to intelligent vision-based video surveillance system that some methods of motion detection and motion alarm are proposed in this paper. At the same time, this paper researches and analyses the theories and techniques of face recognition, and implements them of it.In this paper, an intelligent vision-based video surveillance system is designed. It is based on server/client system model. It includes three parts, such as data acquisition equipment, embedded system-based video server and surveillance center. Embedded system-based video server, which an embedded Linux system runs in it, is the main part.