Palmprint Recognition Using Hessian Matrix and Two-Component Partition Method

Abstract

Palmprint recognition has been comprehensively examined in the past couple of years and various undertakings are done to use it as a biometric methodology for various applications. The point of this study is to construct an effective palmprint recognition technique with low computational multifaceted nature and along these lines to expand the acknowledgment and precision. Since edges are free from distortion, they are very reliable and subsequently used for palm print recognition. The originality of the proposed technique depends on new area of interest (ROI) extraction took after by new principal line extraction and texture matching strategy. The new principal line extraction technique is created by using the Hessian matrix and Eigen value. The texture matching of the ROI is done using new 2-component partition method by segmenting the image into comparative and non-comparative edges. Examinations are finished on a database and exploratory results exhibit that the accuracy of the proposed method is comparable to past methods used for palmprint recognition.