实时人脸识别系统研究

The reseach of the face recognition system based on MATLAB

为研究一种人脸识别的实时检测系统,本文设计了由视频实时采集/预处理模块、数据库管理模块、人脸识别模块和识别信息输出模块组成的识别系统框架。首先利用摄像头采集人脸信息,并通过微软数据库软件SQL储存人脸身份信息,再用MATLAB软件建立GUI人机交互界面,使用经典的主成分分析法PCA提取视频图像中数据的主要特征分量。在MATLAB GUI编辑界面中,通过数据库的链接访问,识别数据库中人物身份信息的操作。通过matlab软件中GUI,建立了人机交互界面,利用SQL Server成功存储录入人脸身份信息,并与人脸录入信息程序实现连接。在MATLAB脚本中对SQL Server实现了已储存人脸信息的查找、读取,录入,删除操作。本系统人脸实时识别率达到90.32%,达到人了脸实时识别效果。

As a real-time detection system of face recognition, this paper designs the framework of recognition system consists of real-time video acquisition/processing module, database management module and face recognition module and output module identification information.first, using the camera face information collected, and through the Microsoft SQL database software stored facial identity information, and then using MAT- LAB software to establish the GUI human-computer interaction interface, using the classical Principal Component Analysis（PCA）finally（Principal Component Analysis, PCA）to extract data in video image, the main characteristics of the components and implementation in MATLAB GUI editor interface, through the link to access the database, identify the identity information of characters in database operation.through the MATLAB software GU1, the establishment of the human-computer interaction interface, the use of SQL Server successfully stored entry face identity information, and with the face input information program to achieve the connection.In the MATLAB script, the stored face information is searched, read, input and deleted by SQL Server.the real-time recognition rate of the system is 90.32%, which can achieve the real-time recognition effect of human face.