DETERRING MALPRACTICE IN A NETWORKED COMPUTER BASED EXAMINATION USING BIOMETRIC CONTROL ATTENDANCE REGISTER

In recent time students are currently adopting various examination malpractice method. The most rampart among the approach is impersonation which cannot be easily detect especially in a very large class and conspiracy of some invigilator or teachers. This research is focused on design of biometric control examination attendance register to deter impersonation during examination. There are various approach to biometric usage which include the fingerprint, face recognition, DNA, hand geometry, iris recognition, retina etc. This research work adopted face recognition biometric technology that recognized different faces. Database of the captured image was built through the use of K-means/hierarchical algorithm model and EM algorithms to initiate and refining the database model respectively. Face recognition was done via skin segmentation, candidates face search, and verification, while face recognition was carried out by face image processing and classification. The entire process was coded using java.net and the resulted system was tested with return shows significant accuracy of recognition test for candidate/students used in the training and testing phase.