On the Use of Time Information at Long Distance in Biometric Online Signature Recognition
In this paper, we investigate the possibility to improve biometric recognition accuracy taking advantage of in-air movements at long distance in online handwriting signature. In-air movements at long distance appear when the distance from the tip of the pen to the paper surface is higher than 1 cm. In this case, the computer can only know the time spent in air because the distance is too high to track the x and y coordinates of the movement. In this paper, we tried several strategies and we checked that the most suitable approach consists of filling the gaps in the acquired signals replicating the last acquired value. In this paper, we check that in air-long movements are three times more frequent in skilled forgeries (1.59%) than in genuine ones (0.48%).
KeywordsOn-line signatures Dynamic time warping In-air movements
This work has been supported by FEDER and MEC, TEC2016-77791-C4-2-R.