Publication

Abstract : Authentication of a person using biometrics has become more popular research work today in the field ofcomputer vision. Face, iris, retina scan, finger print, voice are mostly used for identifying and verifying an individual. In this paper, we present an efficient person identification system using face based on haar features, Bayesian Classifier andHistogram matching algorithm. The Haar features along with Adaptive boosting technique are used to detect and localize the face. The detected face is used for identifying an individual effectively. To recognize the person, we use an imagecomparison algorithm for identifying whether the person is genuine or imposter. The algorithm comprises of two parts, features extraction and features matching. This system can present resultseffectively even if the system has slightocclusions, vary in illumination and different poses of face. We present the experimental results of our algorithm on ourproposed collage database and MUCT database.