Hand-Shape Biometrics Combining the Visible and Short-Wave Infrared Bands

Abstract

This paper proposes a hand shape biometric device with two sensors, respectively working in the visible and 1470nm bands. The inclusion of the 1470nm band sensor is to improve both security and performance. The security is improved by including a spoof detector and the performance by combining both bands. The spoof detector combines three skin detection indices obtained by comparing the reflectance of the hand image in the red, green and blue bands with that from the 1470nm band. The hand tissues reflect the visible radiation while absorbing the 1470nm radiation. The band combination is carried out at a score level which reduces the error rate because different images were obtained under different physical principles (reflection and absorption). The system performance has been evaluated with a database containing 10 acquisitions from each of a group of 100 users and 390 acquisitions from 62 imitated hands made of different materials. The experimental results confirm both security and performance improvement.

A novel hand reconstruction approach and its application to vulnerability assessment

“…The process of aligning the contours can be divided in four stages: i) for each hand image, we automatically locate 14 landmarks (see crosses in Fig. 2) using the methodology proposed in [19]; ii) the contours are aligned by placing the hand geometric center as the coordinate origin, and rotating the hand contour by an angle equal to the slope of the line between the 1st and 3rd finger-web: this allows to reduce the effects of translation and rotation; iii) the envelope line between landmarks is sampled with a number of points equal to average envelope length in GPDS2 DB divided by five; iv) finally, the hand contour is represented as a 2n element vector composed by the coordinates (x and y) of n = 630 selected contour points.…”

“…In this new scenario, the results presented in this contribution show the necessity to include in hand-shape applications efficient countermeasures to repel the studied attacks [19,21].…”

“…• Development: geometry-based system [19]. Geometric features of the hands (48 widths and 4 lengths from the little, ring, middle and index fingers) are obtained by measuring the widths and lengths of each finger.…”

“…-Silhouette-based system [19]. The method employed on this paper is based on direct silhouette alignment of 50 equal spaced samples of the finger contour of the hands excluding the thumb.…”

Hand Geometry Identification System using Machine Vision

“…The number of planes in an image corresponds to the number of arrays of pixels that compose the image. [7] C.…”