Identity recognition method and system based on hand posture, terminal and storage medium

Through the identification method based on hand posture, the encoding and fusion of contour and skeleton features is solved, and the existing biometric identification is easily fraudulent is achieved, achieving higher identification accuracy and security.

CN119723207AActive Publication Date: 2025-03-28GUANGDONG UNIV OF TECH
5 Cites 2 Cited by

Patent Information

Application Number
CN202411911837.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-24
Publication Date
2025-03-28
Estimated Expiration
2044-12-24

AI Technical Summary

Technical Problem

The existing biometric identification methods based on biometrics are easily forged, resulting in the identification methods being easily fraudulent and cannot guarantee the safety of users.

Method used

Using a hand gesture-based identity recognition method, by obtaining the target video and preprocessing each video frame, the contour map and skeleton map are extracted, and the features are encoded using the contour encoding module and the skeleton encoding module. Combining global maximum pooling and global average pooling, it is input to the multimodal spatial fusion module and the multimodal time fusion module to generate spatial feature representations and temporal feature representations, and finally identity identification is performed.

Benefits of technology

By capturing the unique characteristics of hand posture, the accuracy and security of identity recognition are improved, the risk of forgery gestures is reduced, and the problem that existing biometric recognition methods are prone to fraud are effectively solved.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses an identity recognition method and system based on hand postures, a terminal and a storage medium, and the method comprises the steps: obtaining a target video, carrying out the preprocessing of video frames, obtaining a plurality of preprocessed video frames, and obtaining a contour diagram and a skeleton diagram according to a contour extraction network and a skeleton extraction model; encoding each contour diagram to obtain a plurality of contour features, and encoding each skeleton diagram to obtain a plurality of skeleton features; global maximum pooling and global average pooling are used to obtain spatial feature embedding and time feature embedding, the spatial feature embedding is input into a multi-modal spatial fusion module to obtain spatial feature representation, and the time feature embedding is input into a multi-modal time fusion module to obtain time feature representation; and obtaining a classification result according to the spatial feature representation and the time feature representation, and performing identity recognition according to the classification result. According to the invention, accurate hand posture classification for identity recognition can be obtained, and then identity recognition is carried out.
Need to check novelty before this filing date? Find Prior Art