Device unlocking method and device based on facial action

Abstract

The invention provides a device unlocking device based on a facial action, which comprises a facial acquisition unit, a display unit, an input unit, a storage unit, an action classification unit and an image matching unit. Also provided is a device unlocking method based on a facial action, and the method which matches a received facial image of a user with stored facial information of a corresponding user, and judges whether or not verification is passed according to a matching result. The unlocking method based on the facial action provided by the present application improves the safety degree at the time of unlocking, facilitates the user's memory, and refers to the user experience. In addition, the storage space and the transmission bandwidth are greatly saved for the way of storing images by classification.

Description

technical field [0001] The present invention relates to the field of terminal security, in particular to a method and device for unlocking equipment based on facial movements. Background technique [0002] The method of using biological characteristics to identify the current person is to use the biological characteristics of the human body to analyze and authenticate the identity of the person. Traditional biometrics include face, fingerprint, iris of human eye, voiceprint, smell and so on. Biometric identification technology has been widely used in different fields such as personal security, property and financial security, and private information security. [0003] As a face recognition technology in artificial intelligence technology, it has been widely used. With the development of Internet and artificial intelligence technology, face recognition technology has also been further improved. Face recognition technology is mainly used in the unlocking and authentication ...

AI Technical Summary

Problems solved by technology

[0006] Facial action authentication usually uses the captured image as a static password. In the prior art, the image capture device determines whether the features contained in the image information are It is a unique feature of a legitimate user, and then realizes authentication, but there are unsafe factors in this method, for example: by borrowing a photo of a legitimate user as a captured image, it is also possible to pretend to be a legitimate user, and the image capture device cannot distinguish between captured objects It is a photo or a real face, so although the image on the photo has passed the authentication, the person holding the photo is not necessarily a legitimate user, so he can pretend to be a legitimate user and cause harm to the interests of the legitimate user

[0007] The existing technology also includes that when processing captured images, it will be analyzed and processed by algorithms. Since the processing algorithms are relatively mechanized, it is not easy to distinguish people with similar looks. For some special cases such as: two For very similar people, the characteristics analyzed by the system through the algorithm are similar to a certain extent. The system may misjudge that the two people are legitimate users, and then grant wrong permissions to people who are similar to legitimate users, causing security risks

[0008] That is to say, in the prior art, transactions or unlocking are often performed only by identifying the user's frontal face image, and the theft of the face image, or pretending to be the user through photos, models or other advanced means to unlock or There are more and more authentication methods, which also bring security risks to the lives and work of the majority of users

Images