AN EFFICIENT FACE RECOGNITION USING DCT, ADAPTIVE LBP AND GABOR FILTER WITH SINGLE SAMPLE PER CLASS
Nowadays face recognition plays an important role in today’s world. It has achieved greater importance in the
field of information security, law enforcement and surveillance. Now this face recognition approach is applied to many areas
like Airport security, Driver’s License, Passport, Customs and Immigration. In face recognition Local appearance based
methods had achieved greater performance. In this paper we have proposed single sample per class using Discrete Cosine
Transform, Adaptive Local Binary Pattern and Gabor Filter based on local selective feature extraction approach. Discrete
Cosine Transform is used to extract the facial features from the face image .It helps to extract the facial features efficiently.
Then the Gabor filter extracts the textual feature and generates a binary face template based on that features. And this binary
face template act like a mask to extract local texture information using Adaptive Local binary pattern. Adaptive Local binary
pattern method is efficient to face recognition since it is less sensitive to illumination and scaling. And this approach uses
histogram-based matching. It reduces the computational time complexity and space complexity. Here FERET database is
used for face recognition.