Face detection for security surveillance system

This paper explains a new security surveillance system based on face detection. This system uses Haar rectangle as its feature, and selects appropriate classifiers through the AdaBoost learning algorithm for face detection. When it finds face information from video, it will add a label with system time to the face information and save these information into the database. And it provides query and retrieval functions for user. The results show this system can detect human face effectively in some special circumstances, and save the detection information into database after adding a time label. It greatly facilitates the query and retrieval of video surveillance.