Recognition equipment of video identity and recognitionmethod ofvideo identity

Abstract

The invention discloses recognition equipment of video identity and a recognitionmethod ofvideo identity. A computer recognition (image recognition and voice recognition) technology is applied to therecognition of the video identity to form a comprehensive and multi-strategy way combined with deep learning, and the risk to the transaction by the video identity recognition error can be effectivelyprevented. By combining the voice recognition, the identity card face recognition, living body recognition, a traditional identity card and real-name verification, the method mainly comprises the steps of entering identity information, constructing a voice recognition model, constructing a face living body recognition model, online verifying voice recognition, online verifying identity card information and online verifying face recognition and the like, and the recognition equipment and method are the equipment and method capable of more flexibly and reliably recognizing the video identity byutilizing the deep learning technology.

Description

technical field [0001] The present invention relates to the technical field of image recognition, in particular to a method and equipment for video identity recognition Background technique [0002] At present, video identity recognition is mainly through single-strategy speech recognition or face recognition. Face contrast recognition, as a mature biometric technology, has been applied to remittance user registration, user verification and other scenarios. However, face contrast The recognition system can only recognize the faces in the collected images, but cannot judge the authenticity of the collected face images, so it is extremely vulnerable to various deliberate attacks. Once these attacks are successful, the security of user accounts and funds may be threatened, and the consequences and losses may be very serious. In order to solve this problem, face recognition technology emerged as the times require. This technology aims to distinguish whether the face images coll...

AI Technical Summary

Problems solved by technology

[0007] Face recognition technology usually has the following disadvantages: Instability: Compared with other biometrics, such as fingerprints and irises, the stability of the face is not high

Uncontrollability of the acquisition environment: When collecting images, the lighting conditions and the background of the characters are often uncontrollable, and these factors will interfere with the recognition accuracy

Images