A Study of Some New Features for the Palmprint Based Authentication
Madasu Hanmandlu, Ritu Vijay, Neha Mittal · 2011
Abstract—This paper presents an authentication system based on palmprint. The Region of interest (ROI) is extracted from the palmprint image. A tangent is drawn by finding the curves between fingers. The perpendicular bisector of this tangent divides the rectangular area enclosing the palmprint into two equal parts. For the purpose of feature extraction, ROI is divided into a suitable number of non-overlapping windows of different sizes and three types of features, viz. Sigmoid, Energy and Entropy features are extracted. These three sets of features are used for the authentication of users using Euclidean Distance and Support Vector Machine (SVM) as the classifiers. The palmprint as a biometric modality is gaining acceptance in the field of biometrics. As compared to other biometric modalities, it is bestowed with enormous information serving as its discriminating power. We will discuss a few important contributions made on this modality. Matching of palmprints in (5) deals with the feasibility of identifying a person based on a set of features extracted along the prominent palm lines (and the associated line orientation) from a palmprint. Next a decision is made whether two palmprints belong to the same hand by computing a matching score between the corresponding sets of features of the reference and test palmprints. These two sets of features/orientations are matched using point matching technique which takes into account the nonlinear deformations as well as the outliers present in the two sets. An important indexing algorithm is proposed in (10) based on the palmprint classification. This algorithm uses a novel representation involving a two-stage classifier that provides the even-distributed categories. The representation scheme is directly derived from the principal line structures. This scheme does not use wrinkles and singular points and is capable of tolerating poor image quality. In this paper, we will explore features such as Sigmoid, Energy and Entropy extracted from a palmprint. An effort is