Traceable In-Air Signature 3D Restoration Record Structure and In-Air Dominant Hand Biometric Based on Dynamic Time Warping Algorithm

The portable device in-air signature is developing at a fast speed for the past decades because of the advantage of being remote and traceless, but it has gained limited application in the business field. Lacking supervision during the enrollment phase is one of the security risks for in-air signature recognition. We propose a traceable in-air signature structure based on 3D signature restoration for dominant and non-dominant hand in-air signatures. We designed a dominant hand signature recognition model based on the dynamic time warping (DTW) algorithm to reduce the security threat during the enrollment phase of the in-air signature. By applying the DTW model to, in total, 80 in-air signature dominant and non-dominant hand samples collected through the self-designed application, the model can achieve 100% accuracy if more than 6 training samples are taken for each subject and 0% EER for the available participants.