Hand Biometric Segmentation by means of Fuzzy Multiscale Aggregation for Mobile Devices

Biometrics applied to mobile devices is one of the most recent topic of interest in biometrics. Due to the limitations of these devices, in terms of computational cost, biometric techniques must be carefully adapted to this architectures. This paper proposes a quasilinear approach for hand biometric segmentation based on fuzzy multiscale aggregation. The algorithm yields promising results in terms of segmentation accuracy, being tested with hand images acquired with a mobile device in a non-controlled and non-invasive environment. Finally, this approach is compared to the performance of the well-known Normalized Cuts algorithm, with positive results.