User Authentication Using Inner-Wrist Skin Prints: Feasibility and Performance Assessment with Off-the-Shelf Fingerprint Sensor

Wrist-worn devices enable new paradigms of implicit and continuous user authentication; however, identifying biometric modalities that combine reliability with practical integrability remains challenging. Inner-wrist skin texture represents a relatively unexplored biometric characteristic that may be acquired unobtrusively using commodity hardware. This study evaluates biometric verification based on inner-wrist skin texture using an off-the-shelf capacitive fingerprint sensor and an unmodified, manufacturer-provided fingerprint verification algorithm. Two experiments were conducted. Experiment 1 assessed baseline verification performance under controlled acquisition conditions in a cohort of 33 participants (21 male, 12 female; mean age 30.0 ± 16.9 years, range 10-71 years), yielding 1768 genuine authentication trials. Experiment 2 examined the effect of wrist posture variation under controlled flexion in a separate cohort of 15 participants (11 male, 4 female; mean age 30.9 years, range 18-49 years), with 3900 authentication trials recorded. Across 86,897 impostor comparisons in Experiment 1, no false acceptances were observed, corresponding to a conservative upper bound on the false acceptance rate of 6.7 × 10^-5 at the 99.7% confidence level, while the false rejection rate was approximately 2.93%. In Experiment 2, the overall false rejection rate increased to 3.52%, with no clear monotonic relationship between wrist angle and verification performance within the tested range. The results demonstrate that inner-wrist skin texture can be captured and matched using fingerprint-oriented sensing and matching technology under controlled conditions, providing an experimental baseline for this biometric modality. At the same time, the use of a closed matching algorithm and a sensor designed for fingerprints limits interpretability and generalization. These findings motivate further investigation using dedicated recognition methods, larger sensing areas, and extended evaluation protocols tailored specifically to wrist skin print biometrics.