Face Recognition Based Anti-cheat Echnology System Design In Driver Examination

At present, more and more people to get a driver’s license through cheatingmethod with fluky mentality, and this phenomenon will lead to the danger trafficproblem, so the research and development of anti fraud system for driver test isvery important. The popular anti fraud step is to use the fingerprint instrument,but people who will study driving are from different working environment,which limits the fingerprint identification application and popularization. Theface recognition technology becomes a popular means for authentication becauseof its easy to use and no violating. A anti-cheating system in driver testing basedon face recognition is proposed in this paper in order to avoid the driver’scheating acting in their testing. This system can makes for the security of landtraffic. The main contents of the paper are as follows:1. Firstly, the development background of the system is introduced andchallenges and development situation which the important face recognitiontechnology module faces are also discussed.2. The functional requirements, non-functional requirements, and feasibilityproblems of the anti-cheating system based on the analysis of face recognitiontechnology are analyzed. At the same time, the overall system architecture andvarious function modules that should be contained in this system are proposed.3. In detailed design part, image acquisition and image preprocessingmethod are introduced firstly, and the feature extraction and classifier designproblems that are included in face recognition are given the focus of discussion,Subsequently. In many face recognition algorithm, the maximum scatterdifference method is the popular one for its high recognition performance andlow computation complexity. In this paper, the improved algorithm is proposed,and the validity is proved on two popular face databases using MATLAB. At last,the support vector machine is used to classify, and the simulation results basedon libsvm tool box show that the improved maximum scatter difference methodand SVM can be used in this system.