我国铁路客运实名制验票识别技术的探讨

Discussion on Real-name Ticket Checking and Identification Technologies of Railway Passenger Transportation in China

通过阐述自动识别技术在铁路客运实名制验票的应用需求,对指纹识别技术和人脸图像识别技术及其优缺点进行分析,确定选择基于人脸图像识别技术的验票方式,并采用基于概率潜在语义分析模型的人脸身份认证算法,构建人脸词包模型及人脸图像差特征空间,提取人脸标识特征,结合人脸图像标识特征的出现频率,采用支持向量机分类器完成人脸身份认证验票。

Through expounding the application demand of automatic identification technology on real-name ticket checking of railway passenger transportation, this paper analyzes the advantages and disadvantages of the fingerprint identification technology and the face image recognition technology. Based on above, the ticket checking method based on face image recognition technology was determined and selected, and by using the face verification algorithm based on probabilistic latent semantic analysis（p LSA） model, the face model using bag of visual words and the face feature space based on image subtraction were established, and the face identification features were extracted, and combining with the occurrence frequency of the face image identification features, the ticket checking based on face verification was completed by using support vector machine classifier.