Finger Vein Recognition based on Matching Score-Level Fusion of Gabor Features

Abstract

Most methods for fusion-based finger vein recognition were to fuse different features or matching scores from more than one trait to improve performance. To overcome the shortcomings of "the curse of dimensionality" and additional running time in feature extraction, in this paper, we propose a finger vein recognition technology based on matching score-level fusion of a single trait. To enhance the quality of finger vein image, the contrast-limited adaptive histogram equalization (CLAHE) method is utilized and it improves the local contrast of normalized image after ROI detection. Gabor features are then extracted from eight channels based on a bank of Gabor filters. Instead of using the features for the recognition directly, we analyze the contributions of Gabor feature from each channel and apply a weighted matching score-level fusion rule to get the final matching score, which will be used for the last recognition. Experimental results demonstrate the CLAHE method is effective to enhance the finger vein image quality and the proposed matching score-level fusion shows better recognition performance.

Nonintrusive Finger-Vein Recognition System Using NIR Image Sensor and Accuracy Analyses According to Various Factors

“…First, we considered each finger of each person to form a different class. This method is used by conventional finger-vein recognition systems to evaluate the recognition accuracy [ 7 , 8 , 9 , 11 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 25 ]. Consequently, for the good-quality, mid-quality, and open databases, the number of classes were 120 (20 people × 6 fingers), 198 (33 people × 6 fingers), and 636 (106 people × 6 fingers), respectively.…”

“…Work has also been conducted on extracting and combining various features from finger-vein images to increase the quality of the recognition results [ 16 , 17 , 18 , 19 ]. In [ 16 ], they used both the global feature of the moment-invariants method and Gabor filter-based local features.…”

“…In [ 16 ], they used both the global feature of the moment-invariants method and Gabor filter-based local features. In the method proposed by Lu et al [ 17 ], eight-channel Gabor features were extracted and analyzed prior to application to score-level fusion to obtain the final matching score. Qian et al [ 18 ] proposed a finger-vein recognition algorithm based on the fusion of score level moment invariants by the weighted-average method.…”

Improving the Performance of Finger-Vein Recognition System Using A New Scheme of Modified Preprocessing Methods

“…However, the achieved Equal Error Rate (EER) was not promising and their method suffers from the complex computation drawback due to the use of Hough transform for fitting the missing points on the finger border and the implementation of NLM algorithm. Lu and others [15] Used open operation followed by Hough transform for They achieved 98.79% recognition rate using SDUMLA-HMT. In another study, Lu and others [16] respectively using SDUMLA-HMT database.…”

Different Views on the Finger—Score-Level Fusion in Multi-Perspective Finger Vein Recognition

“…They achieved minimum EERs of 2.63 and 0.78%. Lu et al [32] proposed a score-level fusion scheme based on Gabor features. Usually, the individual filter responses obtained from the Gabor filter bank are weighted and/or directly combined into a single output feature.…”