Human authentication with finger textures based on image feature enhancement

Abstract

The main goal of this paper is to authenticate people according to their finger textures. We propose to extract Finger Texture (FT) features of the four finger images (index, middle, ring and little) from a low resolution contactless hand image. Furthermore, we apply a new Image Feature Enhancement (IFE) method to enhance the FTs. The resulting feature image is segmented and a Probabilistic Neural Network (PNN) is employed as an intelligent classifier for recognition. Experimental results illustrate that the proposed approach has superior performance than recent published work. Moreover, the best IFE results were obtained with the Equal Error Rate (EER) equal to 4.07%.

Finger Texture Verification Systems Based on multiple spectrum Lighting Sensors with Four Fusion Levels

“…This rule is usually employed in the FLF and it has been cited that this method has the ability to achieve a significant performance [49]. Interestingly, this method has achieved acceptable EER in the SLF as the input data to the biometric system were doubled and increasing the input information will enhance the biometric system performance as confirmed in [29]. Nonetheless, this rule cannot be considered as a best choice between all the fusion levels, where it has benchmarked inferior EER value in the ScLF.…”

“…Then, the score fusion was executed after the matching operations. In 2015, Al-Nima et al [29] suggested extracting all the FT parts of the four fingers. In this work, it was cited that collecting more FT patterns will lead to increase the recognition performance.…”

“…It can be observed from the previous literature that utilizing FTs for biometric recognition has always been considered as a part of study, except in [29,31]. Also, no publication has concentrated on combining various FT specifications that captured by using different spectra.…”

“…(5) It has been cited that using CV calculations can provide effective descriptions for the variances of the featured images. The resulting images have been partitioned into non-overlapping windows with a fixed size of 5 × 5 and the CV value is computed for each window [20,29,30,33]. To calculate the CV values, the following equations have been exploited [34]:…”

“…2400 images of fingers belong to the WHT spectrum and 2400 images of fingers belong to the BLU spectrum. Firstly, 4000 images were used in the learning stage and 800 images were used in the testing stage as exploited in the state-of-the-art studies [20,[29][30][31]. Secondly, we have changed the number of learning and testing samples (by decreasing the number of learning images and increasing the number of testing images) and we have recorded the obtained evaluations.…”