Palmprint Classification Using Contourlet Transforms and Orthogonal Moments

The increased demand for accurate and reliable solutions for authentication makes Biometrics increasingly gaining importance today. Palmprint is widely considered like fingerprint, for the common features like ridges. It also differs from fingerprint because of its texture based in nature rather than minutiae based. In preprocessing ROI is extracted from the palmprint using a new methodology. Contourlet transform is applied to capture the features before calculating the Zernike moments. Matching is performed by using k-fold cross validation. Since, Contourlet transform is having the property of capturing high dimensional features from multidirectional and Zernike moments have the property of geometrical invariance, they are superior in image representative capability. From the experimental results, it has been observed that Contourlet Transform combined with Zernike moments achieve superior performance than the other well-known models.