多姿态人脸识别技术在安防系统中的应用研究

Research on the Application of Multi-Pose Face Recognition Technology in Security System

为提升安防系统人脸识别的速度及质量,营造更加安全、稳定的防护环境,本文对多姿态人脸识别技术在安防系统中的应用展开研究。首先,进行人脸多重检测,并对识别特征模块化提取,布设多姿态识别节点;其次,构建ORL(Olivetti Research Laboratory)多姿态人脸识别模型,采用支持矢量机(Support Vector Machine,SVM)多维姿态处理人脸识别;最后,本文方法与传统单质人脸识别、传统局部人脸识别进行对比实验。测试结果表明:与传统单质人脸识别测试组、传统局部人脸识别测试组相对比,本文分析的多姿态人脸识别测试组最终得出的识别时间仅为0.42 s,表明其实际识别过程中的速度更快,误差小,具有实际的应用价值。

In order to improve the speed and quality of face recognition in security systems and create a safer and more stable protection environment,this paper investigates the application of multi-pose face recognition technology in security systems.Firstly,we perform multiple face detection and modular extraction of recognition features,and deploy multi-pose recognition nodes;secondly,we construct an ORL(Olivetti Research Laboratory) multi-pose face recognition model,and use Support Vector Machine(SVM) to process face recognition in multi-dimensional pose;finally,our method Finally,this method is compared with traditional singlequality face recognition and traditional local face recognition.The test results show that compared with the traditional single-quality face recognition test group and the traditional local face recognition test group,the final recognition time of the multi-pose face recognition test group analyzed in this paper is only 0.42 s,which indicates that its actual recognition process is faster and has less error,and has practical application value.