Research On Application Of Face Recognition Technology In Examination Systems

As one of the most active biometric identification technology, face recognition technology has been widely applied in various fields such as judicial system, public security, government system, aerospace, education, finance, different kinds of enterprises and institutions, etc. Among all these areas, education is the most typical area that the techonology is applied. In recent years, examination system with its prominent advantages of high security, speediness, accurateness, fairness, justness,low-cost and high-efficiency, has gradually replaced traditional examination and become the mainstream of future examination. At the time examination system has brought us convience, new problems arise, and among them the most crucial one is the security problem of examination system. Recently, in various examinations such as high-school entrance examination, college entrance examination, postgraduate examination, various certificate examination and promotion examination, different levels of cheating behaviors gradually appear. What’s worse, the cheating methods and techonology is becoming more and more advanced and complicated.This study is on the application of face recognition technology in examination system, which is for the strenghtening of identity testing management of the examinees in order to avoid the cheating behaviors. The main researches are as follows:(1)This thesis grasps the present research of facial recognition technology and examination system, analyzes the problem existed and the solutions for the above problems.(2)This thesis analyzes the related technologies of face recognition technology:firstly, coducting face recognition by inputing the images, then extracting typical features and finally confirming the identity based on the identificaiton screening result. The thesis also summarizes the current common face detection methods and face recognition methods.(3)This thesis compares the face recognition method and proposes a skin-colormodeling mehod to extract the outline of the face to get the candidate area. And then the system saves the tested face on the basis of Adaboost method. According to the experiment, this method is more time-saving and efficient.(4)Based on the high dimentionality of human face, using PCA method to reduce dimentionality and remove the relevance of different images. Then the image data are classified by SVM. And finally face recognition is conducted based on the classification result.(5)A solution method which applies the face recognition technology into examination system is proposed on the basis of problem existed. An examination system is developed based on the B/S structure, so the face recognition model is embedded into examination system to test the examinee identity. The feasibility of the methods are tested by the experiment.