Authentication method of writing behavior characteristics based on wrist-worn electronic devices

Abstract

The invention discloses a writing behavior feature authentication method based on a wrist-worn electronic device, which analyzes the user's writing behavior, extracts writing features and finds features with significant discrimination according to acceleration and angular velocity data obtained by accelerometers and gyroscopes , using a single classifier for classification judgment. At the same time, the dynamic time series distance (DTW) between the training data is calculated and compared with the input samples, and the above two judgments are combined to determine whether the user is legal. The collection of behavioral data by the method of the present invention is very convenient, and can be completed by picking up a pen and writing when the user wears smart bracelets, smart watches, etc. It can better reflect the user's writing behavior, and significantly improve the robustness and fault tolerance of identity authentication; and this authentication method has been tested for attacks and is difficult to be attacked.

Description

technical field [0001] The invention relates to the field of device security authentication, in particular to a security identity authentication method for wrist-worn electronic device users. Background technique [0002] With the advancement of social informatization and intelligence, more and more personal information (such as text messages, emails, pictures, etc.) and sensitive information (such as bank account numbers, office data, etc.) are stored in electronic devices. In addition, the frequent occurrence of personal privacy information leakage incidents has gradually brought the security protection of electronic equipment into the public's field of vision. At the same time, with the popularity of electronic wearable devices (smart bracelets, smart watches, etc.), we can use them to perform a more convenient identity authentication process. [0003] Existing identity authentication methods for electronic devices mainly include memory passwords (such as digital passwor...

AI Technical Summary

Problems solved by technology

However, the first method is vulnerable to observation attacks and guessing attacks, so that the unlocking method is stolen by the attacker; authentication methods such as fingerprints and iris require special hardware support; and the second method is not completely safe. The person's fingerprint can be recovered on the computer from the photo of the person waving

Images