The need for robust recognition system in applications of surveillance and biometrics is increasing with the advancements in technology. The conventional method of passwords and pin numbers can be easily hacked,a so these are being rapidly replaced by more secure and reliable biometric methods. This paper produces a new real-time surveillance system in video which makes use of face characteristics of the user for correct identification. The face is first detected using Viola-Jones algorithm, then a hybrid algorithms were used to extract the features and determine the faces linear discriminate analysis and local binary patterns.The proposed approach tested and gives efficient results when compared with other previous approach.