The IUP Journal of Computer Sciences
Texture-Based Palmprint Recognition Using Discrete Wavelet Transformation

Article Details
Pub. Date : Jan, 2022
Product Name : The IUP Journal of Computer Sciences
Product Type : Article
Product Code : IJCS20122
Author Name : Dhulipalla Nagajyothi and K Venkata Ramaiah
Availability : YES
Subject/Domain : Management
Download Format : PDF Format
No. of Pages : 9



Biometric-based recognition is a method of verification that relies on the biological characteristics of each individual. It is processed based on its additional properties of being identical, portable and difficult to reproduce. The paper presents a palmprint recognition system based on texture features using wavelet transformation, for keeping personnel entry records in large enterprises. Using simple approaches-Discrete Wavelet Transformation (DWT) and Standard Deviation (s)-the study extracts some texture characteristics from the ROI images which are then matched using Canberra, Euclidean and Manhattan distance methods. Among all captured images of datasets, K images per user are used to create the training set, with K ranging from 1 to 4. In image recognition mode, they are compared with the remaining images. The true acceptance rate is used to represent the results.


Biometric-based recognition automatically identifies people based on their behavioral or physiological features. Fingerprint, retina, iris, face and palmprint are a few examples of physiological features. Palmprint-based person recognition is a novel biometric technique that is gaining popularity. It is a type of biometric that uses unique features in human palms to recognize human identity. These features include major lines, wrinkles, ridges, minutiae, single points and textures (Aravind and Hemantha, 2012; and Aravind and Vijaya, 2021). Many feature extraction methods for palmprint recognition have already been proposed in the literature (Lunke et al., 2018; and Xiancheng et al., 2020). While these methods proved to be effective, a majority of them carry out feature extraction on each pixel of a palmprint image, while ignoring the image's region property (Li et al., 2002; and Lu et al., 2009). Because of its tiny feature size, fast matching speed and high verification accuracy, palmprint recognition methods are popular


Biometrics, Palmprint recognition, Texture-based recognition, Canberra distance, Euclidian distance, Manhattan distance, Discrete Wavelet Transformation (DWT)