Comprehensive analysis of spectral minutiae for vein pattern recognition

Abstract

Similar to biometric fingerprint recognition, characteristic minutiae points -here end and branch points -can be extracted from skeletonised vein images to distinguish individuals. An approach to extract those vein minutiae and to transform them into a fixed-length, translation and scale invariant representation where rotations can be easily compensated is presented in this study. The proposed solution based on spectral minutiae is evaluated against other comparison strategies on three different datasets of wrist and palm dorsal vein samples. The authors' analysis shows a competitive biometric performance while producing features that are compatible with state-of-the-art template protection systems. In addition, a modified and more distinctive, but not transform or rotation invariant, representation is proposed and evaluated.

A low-cost multimodal biometric sensor to capture finger vein and fingerprint

“…In addition, we also provide a companion protocol for benchmarking the performance of the system (3) A new finger vein recognition method is presented based on a maximum curvature algorithm and computing the spectral minutiae representation of the finger vein pattern. (4) Extensive experiments are carried out on our database, which consistently show the superiority of the proposed finger vein recognition approach when compared against four different well adopted state-of-the-art schemes [7,2,6,4]. Further, the performance achieved using weighted comparison score level fusion of fingerprint and finger vein shows the applicability of the proposed sensor for real-life scenario.…”

“…There exits two variants of spectral minutiae representation schemes, namely: (1) Location based (2) Orientation based. In this work, we employed the location based SMR scheme by considering its improved performance especially on the finger vein biometrics [2]. …”

“…The proposed method differs from the state-of-the-art scheme mentioned in [2] by replacing its multi-scale filter method with the maximum curvature method [7] to accurately extract the finger vein pattern. Given the ROI of the finger vein sample as shown in the Figure 11 (a), we first perform the image enhancement using adaptive histogram equalization [15] as shown in the Figure 11 (b).…”

“…Finally, we use the location of the estimated minutiae points to compute the SMR as illustrated in the Figure 11 (e). The significance of our proposed finger vein recognition scheme over four different state-of-the-art schemes [7,2,6,4] will be discussed in the Section 5. Note that the dimension of the SMR of both finger vein and fingerprint is of same dimension which is 128 × 256.…”

“…In this work, we employed the cross-correlation [2] based comparator to obtain the similarity scores. Let R(m, n) and P (m, n) be the spectral minutiae representation of the reference and probe sample corresponding to the fingerprint (or a finger vein).…”