Robust rotation estimation of slap fingerprint image for E-Commerce authentication

Rotation is one of the most common challenge scenarios in slap fingerprint segmentation. In this paper, we present a robust estimation algorithm of rotation of slap image. Firstly, a new shape descriptor is defined to evaluate the confidence of each candidate foreground region; secondly, the minimum variation is used to estimate the rotation angle based on the confidence. This algorithm has been incorporated in an E-Commerce authentication system based on slap fingerprint identification, ensuring the security and reliability of the E-Commerce trade.