Robust Face Recognition and Tagging in Visual Surveillance System

Face recognition is one of the most successful applications in the field of image processing. Most robust face methods are difficult to improve their performance under certain environments. The paper proposes a method for robust face recognition and implementing a tagging process in video surveillance. It provides an interface for communicating with camera or webcam capturing images storing and sending them along with relevant information. Video based face recognition is much efficient in many real world applications. Motion detection is the first step before the face detection. The face image is divided in to several regions and finding the weight for each region. Face is one of the most suitable biometrics for surveillance applications. It is client server architecture. A user interface is provided at the client and the user interacts with the Web cam view and captures the image. The user sends the relevant data to server for verification. The authentication is based on face matching. After motion detection the face is extracted from the image and calculates the corresponding eigen vectors of each face region. If the face matches with the image in the database the server provides a tagging mechanism. Face is calculated with PCA, LCA, LDA and local binary pattern. There is additional facility provided at the server side to views and generates reports which provide visitor‘s information. The main advantage of this system is that the security in-charge need not be positioned at the secured area. Keywords—Video based image capturing, Face detection, pose and illumination changes, security authentication, Tagging.