A Novel Approach to Palmprint Classification using Orthogonal Moments

Biometrics plays an important role in personal identification and it is becoming increasingly popular today. Palmprint matching is considered in this paper for effective identification of persons. Palmprint matching is performed based on Zernike moments feature descriptors and the classification using Modified Adaboost classifiers. Since, Zernike moments have the property of geometrical invariance, they are superior in image representative capability. From the experimental results, it has been observed that Zernike moments achieve superior performance than the other well-known moments.