Abstract: Utilizing biological features instead of identity cards, the paper proposes an IOT（Internet of Things） based real-name AFC（Automatic Fare Collection） system using technologies of 2.4 GHz RFID（Radio Frequency Identification） and face recognition for wireless perception. Passive RFID tags are used as ticket carriers, and RFID readers are integrated into ticket gates to perceive ticket and passenger information. To reduce the run-time for face detection and recognition with a large number of check-in passengers, an improved Viola-Jones algorithm is proposed to accelerate the face detection, and an improved fisher face method based on OGA（Optimization Genetic Algorithm） with reduced vector dimensionality and optimized feature space is proposed to improve the face recognition efficiency. The experimental results show that the system is reliable for real-name authentification. An example is given to illustrate that the system is feasible for practical applications.

Key words: real-name ticketing, Radio Frequency Identification（RFID）, face recognition, Internet of Things（IOT）

摘要： 以乘客的生物特征取代身份证信息，以2.4 GHz射频识别技术和人脸识别技术实现无线信息感知，建立了基于物联网框架的实名制售检票解决方案。系统采用无源RFID标签为车票载体，通过装有RFID读写器的闸机对车票及其对应的乘客身份信息实现主动感知。为降低现有算法人脸检测及识别的时间，满足大批乘客检票实时性要求，改进了Viola-Jones算法以提高人脸检测速度，提出了基于OGA的改进Fisher脸方法对图像向量进行降维，并对特征空间进行优化以提高人脸识别效率。实验结果表明，设计的实名制票务系统有较好的实时性和可靠性；通过实例说明了系统在实际应用中的可行性。

关键词: 实名制票务, RFID读写器, 人脸识别, 物联网