基于单摄像头的手势识别身份认证方案设计

Design of Gesture Recognition Identity Authentication Scheme based on Single Camera

在新冠病毒肆虐,传统身份认证受到掣肘的背景下,通过对手势识别与控制、图形密码、卷积神经网络等技术的研究,创建一种无接触的身份认证方式。利用单摄像头对手势数据进行采集,同时对手势数据用卷积神经网络训练以提高识别精度生成手势模型,最终通过手势识别对图形密码的选择实现了认证功能。该方案对身份认证方式进行了创新,具有现实的应用价值;随着全息显示技术实用化,该方案会是现实和虚拟之间实用性极佳的认证接口。

In the context of the raging new corona virus and the constraints of traditional identity authentication,a contactless identity authentication method is created through the research of gesture recognition and control,graphic password,convolutional neural network and other technologies.The gesture data is collected by a single camera,meanwhile the gesture data is trained with a convolutional neural network to improve the recognition accuracy and generate a gesture model,and finally the authentication function is realized by the selection of the graphic password through gesture recognition.This scheme innovates the identity authentication method and has practical application value.With the practical application of holographic display technology,the scheme will be an extremely practical authentication interface between reality and virtuality.